MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane Qing Zhang, Xiaoqian Feng, Jing Liu*, Liping Zhao, Xuefeng Song, Peng Zhang**, Lian Gao State Key Lab of Metal Matrix Composites, School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai, 200240, PR China

article info

abstract

Article history:

The methane dry reforming (DRM) simultaneously converts the two greenhouse gases and

Received 19 March 2018

produces syngas (CO þ H2), which is being significant for both environmental and indus-

Received in revised form

trial consideration. Employing well-defined crystal oxides as precursors can produce Ni-

22 April 2018

based DRM catalysts with good sintering and coking resistance by enhancing the metal-

Accepted 3 May 2018

support interactions. Adding basic promoters also is considered as an effective way to

Available online xxx

improve the coking resistance of DRM catalysts, although challenge remains in the control over the structure, morphology and interaction of the promoter in the catalyst. To well

Keywords:

combine the two methods together for better catalytic performance, in this work a Ni/MgO-

Ni nanoparticles

SiO2 catalyst was synthesized through a facile one-pot hydrothermal process, during which

MgO

Ni-phyllosilicate formed as the precursor of Ni particles and MgO promoter was generated

Silica support

in form of Mg-phyllosilicate. This Ni/MgO-SiO2 had a hierarchical hollow sphere structure

Dry reforming of methane

with large surface area (477.4 m2/g). Both the Ni particles (avg. 6.0 nm) and MgO promoter

Hierarchical

uniformly distributed. This hollow hierarchical catalyst performed high activity, thermal stability and coking resistance for catalytic dry reforming of methane. © 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

Introduction Rapid consumption of fossil fuels has triggered a series of environmental and energy crisis. The increasing emissions of carbon dioxide induce severe global warming and climate changes as the result of greenhouse effect. Search for effective methods to realize the reduction or utilization of carbon dioxide is extremely urgent. Methane, the main component of

natural gas, is another kind of greenhouse gas [1,2]. Methane is also considered as an alternative for coal and petroleum as a relative clean energy resource and a chemical source. The industrial utilization of methane has been highlighted recent years due to the worldwide discovery and exploitation of shale gas [3,4]. The dry reforming of methane (Eq. (1)) is an effective route to convert the two greenhouse gases, carbon dioxide and methane into syngas with a low H2/CO ratio, which is suitable for the later production of long chain hydrocarbons

* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (Q. Zhang), [email protected] (X. Feng), [email protected] (J. Liu), [email protected] (L. Zhao), [email protected] (X. Song), [email protected] (P. Zhang), [email protected] (L. Gao). https://doi.org/10.1016/j.ijhydene.2018.05.010 0360-3199/© 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

2

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

through Fischer-Tropsch synthesis and oxygenates through oxo-synthesis [5e8]. This reaction simultaneously consumes these two destructive greenhouse gas and produces valuable chemical products, being of great benefit to both environmental and commercial consideration [9e11]. CH4 þ CO2 / 2H2 þ 2CO, DH298 ¼ þ247 kJ/mol

(1)

Stable and economical catalysts are in urgent need to realize the industrialization of methane dry reforming. Traditional noble metal (Rh, Ru, Pt, and Pd) catalysts present high catalytic activity, superior resistance to carbon deposition and sintering in DRM [12]. Nevertheless, considering the exorbitant price and limited reserves, noble metals are not industrial competitive with other transition metals. Nickel based catalysts are the most favorable catalysts for DRM in recent years for its high activity, abundant reserves and bargain price [13e17]. However, Ni based catalysts usually undergo severe deactivation due to sintering and coke deposition. Many researches were carried out to solve these two problems, especially the carbon deposition which plays the main role in catalysts inactivation [18e22]. Recent years, many researchers employed well-defined crystal oxides as nickel precursors (e.g. perovskites ABO3 [23,24]，spinels ABO4 and A2BO4 [25,26]，hexaaluminates AByAl12-yO19-d [27], phyllosilicates [28,29] and solid solutions [30,31]；B¼Ni) to obtain well-dispersed small Ni particles with strong metal-support interactions [32]. Ni species are homogeneously dispersed in such defined structures. The crystal lattice effectively limits the growth and migration of Ni species, thus Ni nanoparticles with uniform distribution could be obtained after reduction. As a result, catalysts prepared from these precursors usually result in improved catalytic activity and thermal stability [32e34]. In our previous work [25], we found that Ni/g-Al2O3 catalyst reduced from NiAl2O4 performed superior sintering and coking resistance than those reduced from NiO/g-Al2O3. Addition of basic promoters (e.g. K2O [35,36], CaO [37,38], MgO [29,39,40], La2O3 [41], CeO2 [42,43] and Ga2O3 [44]) is another effective method to improve the catalyst activity and coking resistance of DRM catalysts through increasing the basicity of supports. It is generally accepted that supports with basic properties are helpful in adsorption and activation of CO2 species on the catalysts, suppressing the carbon formation through CO disproportionation [13,44,45]. Among these promoters, MgO was reported to exhibit superior catalytic performance and coke suppression [46]. The bargain price of MgO, as well as nickel, contributes to its competitiveness for a future industrial application in DRM. MgO promoter is also valid in promoting catalyst activity and stability through enhancing the sintering resistance and metalsupport interaction [20]. Guo et al. found that the formation of MgAl2O4 phases in Ni/MgO-g-Al2O3 could suppress the formation of NiAl2O4 and stabilize the Ni nanoparticles, leading to higher catalytic activity, coking resistance and sintering resistance [20]. However, MgO promoter was generally added through traditional wet impregnation method or co-precipitation method, which resulted in a weak interaction and inhomogeneous distribution on the support. The weak promoter-support interaction may cause aggregation of the

promoter and coverage of metal active sites by the promoter. Therefore, the challenge still remains in the control over the structure, morphology and interaction of the MgO promoter in the catalysts. Herein we designed and synthesized a MgO promoted Ni/ MgO-SiO2 catalyst. Ni and Mg elements were introduced onto the SiO2 support simultaneously by a one-pot hydrothermal method to guaranty both stronger metal-support and promoter-support interactions. The stronger metal-support interaction could be achieved from reduction of the welldefined Ni-phyllosilicate precursor. The homogeneous distribution of MgO during hydrothermal deposition is desired for the enhanced promotion effect to the catalysis. The growth process of the hollow hierarchical Ni/MgO-SiO2 nanocomposites during hydrothermal condition and hydrogen reduction was investigated. Its catalytic performance and coking resistance for DMR were compared with Ni/SiO2, IMNi/SiO2, and IM-Ni/MgO-SiO2 to investigate the effects of preparation methods, hierarchical structure, and MgO addition. The structural stability of the catalysts at high reaction temperatures was also compared and discussed.

Materials and methods Materials Ethyl alcohol (C2H5OH), isopropanol (C3H7OH), Tetraethyl orthosilicate (TEOS, 99%), NH3$H2O (28%), CO(NH2)2, Ni(NO3)2$6H2O and Mg(NO3)2$6H2O were obtained from Sinopharm Chemical Reagent Co. Ltd, China. All the reagents were used without additional purification.

Synthesis Synthesis of SiO2 spheres SiO2 spheres were synthesized through Stober method. TEOS (1.0 mL) was slowly dropped into the mixture of deionized water (15 mL) and isopropanol (35 mL) under magnetic stirring. Then aqueous solution of ammonia (1 mL) was dropped into the solution. After stirring for 2 h at room temperature, the precipitated silica spheres were collected by centrifugation and washed 3 times with ethanol and deionized water.

Synthesis of Ni/MgO-SiO2 and Ni/SiO2 SiO2 spheres (0.12 g, equivalent to 2  103 mol), Mg(NO3)2$6H2O (0.26 g, equivalently 1  103 mol) and CO(NH2)2 (2.70 g, equivalently 0.045 mol) were dispersed in the mixture of deionized water (8 mL) and ethyl alcohol (9 mL) under sonication and aqueous solution of Ni(NO3)2$6H2O (8 mL, 0.1 M) was dropped in. After 5 h, the suspension was transferred into a 50 mL Teflon-lined stainless steel autoclave and heated at 190  C for 36 h. The resulting precipitates were collected by centrifugation and washed 3 times with deionized water. The hydrothermal product was dried under vacuum at 60  C for 10 h and then reduced in 5% H2/Ar at 700  C for 5 h. The reduced product was denoted as Ni/MgO-SiO2. To investigate the effect of MgO promoter, Ni/SiO2 was synthesized as a reference sample through the same hydrothermal method using SiO2 spheres (0.12 g), CO(NH2)2 (2.70 g)

Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

and Ni(NO3)2$6H2O solution (8 mL, 0.1 M). The hydrothermal product was dried and reduced via the same procedures that were carried out for synthesis of Ni/MgO-SiO2.

Synthesis of IM-Ni/MgO-SiO2 and IM-Ni/SiO2 IM-Ni/MgO-SiO2 and IM-Ni/SiO2 were synthesized through impregnation method. In the preparation of IM-Ni/MgO-SiO2, SiO2 spheres (0.12 g) and Mg(NO3)2$6H2O (2.70 g) were dispersed in Ni(NO3)2$6H2O solution (8 mL, 0.1 M) under sonication followed by mechanical agitation at 60  C until entirely evaporating of liquid. IM-Ni/SiO2 was prepared under the same condition from SiO2 spheres and Ni(NO3)2$6H2O solution. The samples were dried at 100  C under vacuum and reduced at 700  C for 5 h to obtain IM-Ni/MgO-SiO2 and IM-Ni/ SiO2.

3

the complete decomposition of nitrate and avoid interference to the test. CO2 temperature-programmed desorption (TPD) was carried out to measure the basic sites of the catalysts. Reduced catalysts (100 mg) were fixed in a quartz U-tube reactor and pretreated in Ar atmosphere at 200  C for 1 h. After cooling down to 30  C, CO2 was fed into the reactor at a rate of 20 mL/min for 1 h. Then the sample was purged with Ar for 30 min to eliminate physically adsorbed CO2 and heated to 950  C at a rate of 10  C/min to acquire CO2 desorption curve. The carbon deposition on spent catalysts was determined by TGA (thermo gravimetric analysis), DTG (derivative thermo gravimetry) and DSC (differential scanning calorimetry) curves obtained from a simultaneous thermal analyzer (SDT/ STD-Q600, US). The measurements were carried out in a flow of air of 100 mL/min and at a heating rate of 10  C/min from room temperature to 900  C.

Material characterization Catalytic performance The XRD patterns of samples were identified by an X-ray diffraction (XRD, LabX XRD-6100, SHIMADZU, Japan) using a Cu Ka source (l ¼ 1.542
A), with a scan step of 0.02 and a scan  speed of 5 /min between 10 and 90 . The crystalline sizes of the Ni nanoparticles were estimated using Scherrer's equation (Eq. (2)). D¼

Bl b1=2 cos q

(2)

The microstructure and surface morphology of the fresh and spent catalysts were characterized through a scanning electron microscopy (SEM, FE-SEM Mira3/MIRA3, TESCAN, Chech). Elemental composition of Ni/MgO-SiO2 catalyst was investigated though an Energy Dispersive Spectroscopy (EDS, XFlash6160) on the SEM. The morphologies and particle size distribution of catalysts were investigated through a transmission electron microscope (TEM, Tecnai G2 spirit Biotwin, FEI, US) with an acceleration voltage of 120 kV. The element contents of Ni and Mg in the catalysts were analyzed by an inductive coupled plasma emission spectrometer (ICP, 7500A, Agilent Technologies, US). XPS spectrometer (AXIS Ultra DLD, Kratos, Japan) was employed to determine the surface composition and chemical binding energy of element Ni and Mg on Ni/MgO-SiO2. The specific surface areas and pore size distribution of the catalysts were determined by N2 adsorption/desorption isotherm at 77K using a Physisorption and Chemisorption Analyzer (AutosorbiQ-C, Quantachrome, US) in the P/P0 range of 0.04e0.99. The surface areas and pore size distribution were calculated from the adsorption isotherm curves using Brunauer-Emmett-Teller (BET) method and density functional theory (DFT) method, respectively. Hydrogen temperatureprogrammed reduction (TPR) was conducted on Quantachrome Autosorb IQ-C instrument with a thermal conductivity detector (TCD). In a typical process, 50 mg of a sample was fixed by silica wool in an aquartz U-tube reactor. Mixture of H2 and N2 (1/10, v/v) was fed into the reactor at a rate of 40 mL/ min and heated from room temperature to 900  C at a heating rate of 10  C/min. The hydrothermal samples were used after drying at 100  C under vacuum. The impregnation samples were calcined at 400  C for 2 h before TPR analysis to insure

The DRM activities of the catalysts were determined by GCTCD (GC7900, Tianmei, China). 100 mg of Ni/MgO-SiO2 catalyst (20e40 mesh) was installed in a quartz tube reactor fixed on a continuous fixed-bed flow reactor (Fantai-4100, Fantai, China). Before dry reforming reaction, pre-reduction was taken in 5% H2/Ar (20 mL/min) at 700  C for 1 h. After reduction, the reactant with a molar ratio of 1:1:1 (33.3% CH4, 33.3% CO2, 33.3% Ar; flow rate ¼ 30 mL/min, GHSV ¼ 18,000 mL/ (h$gcat)) was induced in the flow reactor. The catalytic activities in DRM of the catalysts were evaluated under a temperature range of 500  Ce700  C. The catalytic stability of the catalysts was evaluated under 700  C for 50 h. The effluent gases from the reactor were analyzed on-line by gas chromatography with a thermal conductivity detector (TCD). CO2 and CH4 conversion of the catalysts were calculated through the gas flow rate by Eq. (3) and Eq. (4), respectively. CH4 Conversion : XCH4 ð%Þ ¼

FðCH4 Þin  FðCH4 Þout  100% FðCH4 Þin

(3)

CO2 Conversion : XCO2 ð%Þ ¼

FðCO2 Þin  FðCO2 Þout  100% FðCO2 Þin

(4)

Catalytic stability of the catalysts in DRM was investigated through a 50-h test under 700  C.

Results and discussion Structure and morphology The XRD patterns of the Ni/MgO-SiO2 catalyst after hydrothermal reaction and hydrogen reduction were compared in Fig. 1. The diffraction peaks (2q ¼ 19.4 , 26.7 , 33.7 , 39.7 , 41.0 , 43.3 , 53.2 , 55.7 , 60.9 , 71.4 and 73.3 ) in Fig. 1A were indexed to the nickel phyllosilicate Ni3Si4O10(OH)2$5H2O (JCPDS 430664) and the diffraction peaks (2q ¼ 19.4 , 28.8 , 34.5 , 36.2 , 53.2 ,60.9 ) were indexed to magnesium phyllosilicate Mg3Si4O10(OH)2 (JCPDS 03-0174) [47,48]. The XRD patterns confirmed that Ni2þand Mg2þ existed in the form of phyllosilicates rather than bulk NiO and MgO, suggesting strong

Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

4

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

Fig. 1 e XRD patterns of Ni/MgO-SiO2 catalyst before (A) and after (B) reduction.

interactions between metal species and the support SiO2. Besides the phyllosilicates, the rest of SiO2 existed in the form of non-crystalline phase (JCPDS 29-0085) at 2q ¼ 22.0 [14]. After hydrogen reduction, Ni metal (JCPDS 04-0850) nanoparticles were from exsolution of Ni-phyllosilicate precursor, as shown by the diffraction peaks at 2q ¼ 44.5 , 51.8 and 76.4 . The crystallite size of Ni nanoparticles based on the full-width at half-maximum (FWHM) of Ni (111) plane was approximately 9 nm calculated by Scherrer's equation (Table 1). The XRD patterns of the reduced Ni-SiO2, IM-Ni/MgO-SiO2 and IM-Ni/ SiO2 catalyst were also investigated (Fig. S1). Diffraction peaks were identified to be metallic Ni (JCPDS 04-0850) and noncrystalline phase SiO2 (JCPDS 29-0085) in Ni/SiO2 (Fig. S1A) and IM-Ni/SiO2 (Fig. S1C). The Mg2þ in IM-Ni/MgO-SiO2 existed in the form of bulk MgO (JCPDS 45-0946) at 2q ¼ 36.9 , 42.9 , 62.3 , 74.7 and 78.6 instead of magnesium phyllosilicate (Fig. S1B). The crystallite size of Ni nanoparticles in the controlling samples were also estimated through Scherrer's equation. The average particle size was 11 nm in Ni/SiO2, 21 nm in IM-Ni/MgO-SiO2 and 18 nm in IM-Ni/SiO2 (Table 1). Element contents of Ni and Mg species in the reduced catalysts were identified through ICP, which were close to the theoretical content as shown in Table 2. XPS spectra were employed to investigate the surface composition and the chemical states of the elements on the surface of Ni/MgO-SiO2. The Ni 2p and Mg 1s XPS spectra before and after reduction are shown in Fig. S2. The binding energies of Ni 2p3/2 and Ni 2p1/2 are respectively 857.6 eV and 875.3 eV in the hydrothermal product, which is consistent with the nickel

Table 1 e Ni particle size in the catalysts before and after catalysis. Sample Ni/MgO-SiO2 Ni/SiO2 IM-Ni/MgO-SiO2 IM-Ni/SiO2

DXRD (nm)

DTEM (nm)

D#XRD (nm)

D#TEM (nm)

9 11 21 18

6.6 8.6 10e50 10e100

13 14 19 19

8.3 12.5 10e100 10e100

DXRD: Ni particle size detected through XRD; DTEM: Ni particle size detected through TEM. D#XRD: Ni particle size detected through XRD after 50 h catalysis; D#TEM: Ni particle size detected through TEM after 50 h catalysis.

Table 2 e Element percent and structural properties of the catalysts. Sample

Ni wt%

Ni wt%*

Mg wt%

Mg wt%*

A (m2/g)

A# (m2/g)

Ni/MgO-SiO2 Ni/SiO2 IM-Ni/MgO-SiO2 IM-Ni/SiO2

18.4 25.8 19.1 26.4

22.7 28.1 22.7 28.1

9.8 0 11.1 0

11.6 0 11.6 0

477.4 217.8 39.4 30.5

387.3 102.4 50.9 55.3

M wt%: metal weight percent detected through ICP; M wt%*: Theoretical metal weight percent calculated from the molar ratio of reagents. A: surface area; A#: surface area after 50 h catalysis at 700  C.

silicate [49,50]. The binding energy of Mg 1s is 1305.7 eV, which is also consistent with the magnesium silicate [51]. The XPS spectra further indicate the formation of phyllosilicates after hydrothermal reaction as the XRD pattern (Fig. 1A) revealed. After hydrogen reduction, the binding energy peaks of Ni 2p3/2 and Ni 2p1/2 move to 853.0 eV and 873.5 eV. This suggests that Ni species were from the exsolution of Ni-phyllosilicates. The binding energy peaks are considered to consist of metallic Ni0 (Ni 2p3/2 854.9eV) and NiO (Ni 2p3/2 856.6eV). No NiO diffraction was observed from the XRD pattern (Fig. 1B). However, an inevitable surface oxidation occurred when the sample was exposed to the air before XPS detection was carried out. As a result, the Ni species existed in the mixture of metallic Ni and NiO in the top 2e5 nm layer of the catalyst. Therefore, an insitu pre-reduction before catalytic test is essential to eliminate the surface oxidation and ensure the formation of metallic Ni on the surface. Only slight shift was observed in the binding energy of Mg 1s (1304.9 eV) after reduction, indicating the reservation of magnesium silicate, rather than the formation of bulk MgO (Mg 1s 1303.9 eV [51]). The surface morphology of Ni/MgO-SiO2 was characterized by SEM (Fig. 2A and B). From the SEM images, Ni/MgO-SiO2 was consisted of spheres with a diameter of around 300 nm. According to the XRD results in Fig. 1, the spheres were composed by (Ni, Mg)3Si4O10(OH)2 and SiO2. Ni and Mg species were disperse uniformly on SiO2 support as shown in the EDS results (Fig. S3). The outer surfaces of the spheres were covered by interconnected nanoflakes which construct porous surface structure. The inside of the spheres was hollow as

Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

5

Fig. 2 e SEM images of Ni/MgO-SiO2 catalyst (A) before and (B) after hydrogen reduction; TEM image of Ni/MgO-SiO2 catalyst (C, D) before and (E, F) after hydrogen reduction.

observed from some cracked spheres. After hydrogen reduction, nanoparticles were observed on the surface of the spheres, which was attributed to the formation of metallic Ni (Fig. 2B). The spheres and the porous surface maintained the same after reduction. According to the result of N2 adsorption/ desorption measurement, the surface area of Ni/MgO-SiO2 was 477.359 m2/g (Table 2) and the pore size distribution exhibited a maximum centered at 3.8 nm (Fig. S4B). The nitrogen adsorption/desorption isotherm of the reduced Ni/ MgO-SiO2 catalyst (Fig. S4A) presented a typical type IV curve indicating a mesoporous structure with a H2 hysteresis loop resulting from the cylindrical channels among particles [52]. The large surface area and porous surface structure facilitated the transport and contact of reaction gases. Transmission electron microscopy (TEM) was employed to obtain more information of the hollow sphere structure (Fig. 2CeF). TEM images showed that the diameter of the hierarchical spheres was about 300 nm and the thickness of the shell wall was ~5 nm. Phyllosilicate nanoflakes covered on the outside shells of the spheres present a scale dimension of 20e50 nm. After reduction, no sintering or shape deformation of hierarchical sphere was observed, suggesting a good structural stability (Fig. 2C and D). Nickel species was reduced from the nickel silicate composite after hydrogen reduction

and the particle size distribution was investigated. The Ni nanoparticles were dispersed uniformly on the hollow sphere with an average diameter of 6.6 nm. Very few particles over 10 nm were observed (Fig. 2F). The high dispersion and small size of Ni nanoparticles were attributed to the strong metalsupport interactions obtained from well-defined Ni-phyllosilicates precursor [20,53]. Restriction from Mg-phyllosilicate crystal lattice also inhibits migration and growth of Ni species. Good dispersion of active metal and small particle size were promising for high catalytic activity and coking resistance [12]. To investigate the effect of MgO addition on the hierarchical sphere, we compared the SEM and TEM images of Ni/ SiO2 catalyst with Ni/MgO-SiO2 (Fig. 3). Ni/SiO2 presented a similar structure hierarchical hollow spheres with porous outside surface as Ni/MgO-SiO2. However, spheres with relatively smooth outside surfaces were observe (Fig. 3A). Some solid spheres or spheres with low hollowness degree were also observed in TEM images (Fig. 3B). The incomplete hollowness phenomenon led to a surface area of 217.8 m2/g, which was significantly lower than the surface area of the standard sample Ni/MgO-SiO2 (477.4 m2/g). From the TEM images, most of the Ni nanoparticles in reduced Ni/SiO2 were 6e13 nm and the average size was 8.6 nm. The only difference between the

Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

6

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

Fig. 3 e (A) SEM and (B, C) TEM images of Ni/SiO2 catalyst after hydrogen reduction. preparation route of Ni/MgO-SiO2 and Ni/SiO2 is the presence of Mg2þ in the precursors of Ni/MgO-SiO2. Therefore, the growth mechanism of the hollow sphere structure could be deduced by combining two kinds of hydrothermal samples and the results of our previous research [50]. Firstly, SiO2 solid spheres partly dissolved in alkaline solution and react with metal ions Ni2þ and Mg2þ to form layered phyllosilicates sheets. The phyllosilicates then deposited on the outside of SiO2 spheres in the form of nanoflakes. (Fig. 4). With the hydrothermal reaction continued, the inner silica was dissolved and redeposited completely and a hierarchical hollow sphere structure was obtained. The addition of Mg2þ facilitates the formation of more phyllosilicate sheets, promoting the dissolution of the SiO2. As a result, the hollowness degree in Ni/MgO-SiO2 was higher and more thorough than that in Ni/ SiO2 and corresponded to a larger surface area. To investigate the effect of synthesis route and Niphyllosilicate precursor on morphology of the catalysts, we also investigated the microstructures of the impregnated control samples, IM-Ni/MgO-SiO2 and IM-Ni/SiO2 (Fig. S5). IMNi/MgO-SiO2 was composed by SiO2 solid spheres with a surface area of 39.4 m2/g (Fig. S5A-C). SEM image revealed the aggregation and heterogeneous distribution of Ni nanoparticles and MgO on the surface of SiO2. Although Mg2þ was added as well, the impregnated Mg2þ in IM-Ni/MgO-SiO2 existed in form of bulk MgO rather than Mg-phyllosilicate (Fig. S1). Ni nanoparticles were partially covered by the aggregated MgO and it was difficult to investigate the size distribution through TEM images. However, Ni particles larger than 50 nm were observed in Fig. S3C, confirmed the severe sintering in IM-Ni/MgO-SiO2. IM-Ni/SiO2 catalysts was composed by SiO2 solid spheres with a surface area of 30.5 m2/ g (Fig. S5D-F). Bulk Ni particles obtained through impregnation, which presented a wide size distribution between 10 nm and 100 nm, were observed on the solid SiO2 sphere. The large Ni metal particles in impregnated samples suggested weak

metal-support interactions in IM-Ni/MgO-SiO2 and IM-Ni/ SiO2. Ni particles aggregated rapidly to form large particles as a result of Ostwald ripening during reduction. The morphology comparison of IM-Ni/MgO-SiO2 and IM-Ni/SiO2 further indicated that MgO was coated on SiO2 solid spheres in IM-Ni/MgO-SiO2, which slightly increases the surface area.

Catalytic performance Effect of MgO on catalytic activity The effect on catalytic activity of MgO addition in the form of well dispersed Mg-phyllosilicate was investigated through the comparison between Ni/MgO-SiO2 and Ni/SiO2 catalysts, which possessed similar morphology and structures. CH4 and CO2 conversions of the catalysts at a temperature range from 500  C to 700  C for DRM were calculated and drawn as a function of temperature in Fig. 5A and B. It was obvious that the CH4 and CO2 conversions increase with reaction temperature for all the tested catalysts. Under investigated temperatures, the calculated CH4 conventions of the catalysts were lower than CO2 conversions due to the accompanying reverse water-gas shift reaction (RWGS, Eq. (5)) [1]. The H2/CO ratios of the gaseous products were less than 1, which also indicates the occurrence of reverse water-gas shift reaction (Fig. S6A). Under all the test temperatures from 500  C to 700  C, Ni/MgO-SiO2 presented a superior conversion for both CH4 and CO2. The CH4 conversion were 88% for Ni/MgOSiO2 and 72% for Ni/SiO2 under 700  C. Meanwhile the CO2 conversion were 92% for Ni/MgO-SiO2 and 83% for Ni/SiO2 under 700  C. Considering the similar hierarchical structure of Ni/MgO-SiO2 and Ni/SiO2, it seemed that the addition of MgO played a significant role in the catalytic activity. On one hand, the basicity of MgO promoter was reported to enhance the chemisorption and activation of CO2 and accelerate the reaction [39,45,54,55]. On the other hand, the added MgO promoter combined tightly with SiO2 to form Mg-

Fig. 4 e The growth process of Ni/MgO-SiO2 nanocomposites during hydrothermal condition and hydrogen reduction. Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

7

Fig. 5 e The catalytic activities in DRM of the reduce catalysts: (A) CH4 conversion and (B) CO2 conversion under a temperature range of 500  Ce700  C.

phyllosilicate Mg3Si4O10(OH)2, which provided a channel for carrier activation of CO2 in Ni/MgO-SiO2. There was a strong agreement that CO2 activates on the surface of both active metal and acidic/basic supports [12]. For catalysts with inert support like SiO2, only metal-surface activation could be achieved [56]. On the surface of acidic supports, CO2 can be activated through formation of formates with the surface hydroxyls; whilst on the surface of basic supports, the oxycarbonates can be generated [57e59]. According to the XRD patterns, the support in Ni/MgO-SiO2 was a mixture of and the basic Mg-phyllosilicate amorphous SiO2 Mg3Si4O10(OH)2; whilst the support in Ni/SiO2 was merely amorphous SiO2. Therefore, CO2 activation in Ni/MgO-SiO2 could be achieved through both metal activation and Mgphyllosilicate support activation. Thus the adsorption and activation rate of CO2 were accelerated, leading to superior catalytic activity of Ni/MgO-SiO2 at all the investigated temperatures (Fig. 5). The catalytic activity comparison between Ni/MgO-SiO2 and the catalysts reported in other papers with similar test conditions was shown in Table S1. Ni/MgO-SiO2 expressed a remarkable catalytic activity compared with other catalysts under a relatively lower temperature (700  C). H2þCO2/H2O þ CO, DH298 ¼ þ41.2 kJ/mol

(5)

The CO2-TPD results, Fig. S7, can help to indicate the contribution of MgO in CO2 absorption. The CO2 desorption curve can be divided into three regions. The small desorption peak in Regions I is due to CO2 molecules absorbed in weak basic sites and the remnant of physically absorbed CO2; the broad peak in Region II is attributed to medium basic sites; whilst the peak in Region III is resulted from the CO2 that strongly interacted with surface basic sites. For the Ni/SiO2 and IM-Ni/SiO2 samples, no peaks were observed in Region III, indicating lower CO2 absorbability. Both Ni/MgO-SiO2 and IMNi/MgO-SiO2 yielded peaks in Region III, suggesting stronger CO2 absorbability; the former showed a more intense peak than the later, probably because the Ni/MgO-SiO2 sample has larger surface area and more uniform distribution of MgO. It should be noted that the magnesium silicate decomposed at ~870  C, which could attribute to part of the intense peak of the Ni/MgO-SiO2 in Region III.

Effect of synthesis route on catalytic activity The effect of hydrothermal route and well-defined Ni-phyllosilicate precursor was revealed through comparing with IMNi/MgO-SiO2 catalyst, which possessed the same chemical composition with Ni/MgO-SiO2. Apparently, Ni/MgO-SiO2 exhibited a superior catalytic activity under the investigated temperature. The CH4 and CO2 conversions were 72% and 84% for IM-Ni/MgO-SiO2 under 700  C. The significant difference in the catalytic activity between Ni/MgO-SiO2 and IM-Ni/MgOSiO2 was ascribed to their microstructures and metal-support interactions. The high specific surface area (477.4 m2/g) and hierarchical hollow structure in Ni/MgO-SiO2 promoted the adsorption, diffusion and contact of the reactant gases, which greatly benefited the catalytic efficiency. MgO in Ni/MgO-SiO2 was dispersed uniformly in form of Mg3Si4O10(OH)2 as EDS result suggests (Fig S3). The high dispersity of MgO provided a channel for carrier activation of CO2 through Mgphyllosilicate, which facilitates the catalytic reactions. In contrast, in the impregnated sample IM-Ni/MgO-SiO2, the SiO2 solid sphere structure with a small specific surface area (39.4 m2/g) hindered gas diffusion. Considering the inert nature of SiO2 support, reactant gas could merely be adsorbed and activated on the highly calcined Ni metal particles and the aggregated MgO. The catalyst preparation methods also have critical impact on the interaction between Ni metal particles and the support, which played a significant role in catalytic performance. It was widely reported that catalysts synthesized through conventional impregnation method usually leads to weak metalsupport interaction [60,61]. Formation of well-defined crystal oxide precursors before reduction was an effective way to obtain Ni metal particles with strong metal and support interactions (SMSIs) [31e33]. In hydrothermal sample Ni/MgOSiO2, Ni metal nanoparticles were obtained from hydrogen reduction of well-defined crystal Ni-phyllosilicate Ni3Si4O10(OH)2$5H2O, in which Ni2þ species contact tightly with SiO2. The resulting Ni present well dispersed small particle (~6 nm) as the TEM images show (Fig. 2), thus providing sufficient accessible metal surface and active sites for the catalytic reaction. In IM-Ni/MgO-SiO2, Ni metal nanoparticles were reduced from impregnated Ni2þ species with weak interaction to SiO2. The obtained Ni nanoparticles in

Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

8

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

impregnated samples suffered from severe sintering after hydrogen reduction (Fig. S5). The sintering of active metal particles reduced the accessible metal surface and decrease the activity sites. The disparity in CH4 and CO2 conversions between Ni/SiO2 and IM-Ni/SiO2 further indicated the advantage of hydrothermal method and the hierarchical hollow structure. The CH4 and CO2 conversions were 72% and 83% for the former, 57% and 66% for the latter under 700  C. Even if no MgO was added, the hydrothermal product Ni/SiO2 also expresses a higher catalytic activity to impregnated product under test conditions. It was widely recognized that TPR profiles could reflect the interaction between metal species and the support [62,63]. TPR analysis was employed to compare the metal-support interaction in Ni/MgO-SiO2 and the control samples (Fig. 6). In the TPR profile of Ni/MgO-SiO2, there was a small H2 adsorption peak centered at 505  C and a broad adsorption peak from 600  C to 900  C with a center at around 735  C. A similar TPR curve was observed in Ni/SiO2, where the two main reduction peaks located at around 544  C and 748  C. In these two hydrothermal samples, the weak adsorption peak at lower temperature was attributed to little amount of Ni species existing in the form of bulk NiO [42], which presents relatively weak interaction to SiO2 support. The intense broad

Fig. 6 e TPR profiles of (a) Ni/MgO-SiO2, (b) Ni/SiO2, (c) IMNi/MgO-SiO2 and (d) IM-Ni/SiO2.

peak cantered at 750  C was the reduction peak of Niphyllosilicate Ni3Si4O10(OH)2$5H2O with strong interaction to the support, which was determined by XRD profile (Fig. 1A). The low-temperature reduction peak in Ni/SiO2 was broader than that in Ni/MgO-SiO2, which meant a higher ratio of Ni species with relatively weak metal-support interaction. Moreover, a small peak at around 867  C was also observed merely in Ni/MgO-SiO2, which can be attributed to the decomposition of Mg-phyllosilicate. In the impregnation IMNi/MgO-SiO2 and IM-Ni/SiO2 samples, a sharp intense reduction peak was observed at 348  C and 385  C, respectively. The reduction temperatures in IM-Ni/MgO-SiO2 and IM-Ni/SiO2 were much lower than that in Ni/MgO-SiO2 and Ni/SiO2. The TPR profiles indicated that the metal-support interactions in hydrothermal samples were stronger than that in the impregnated samples, consistent with the XRD (Fig. 1, Fig. S1) and TEM results (Figs. 2e3, Fig. S5).

Catalytic stability, coking resistance and thermal stability Among the tested catalysts, Ni/MgO-SiO2, Ni/SiO2 and IM-Ni/ MgO-SiO2 kept stable catalytic activities during the whole test (Fig. 7), whereas the CH4 and CO2 conversions of IM-Ni/SiO2 degraded gradually with time. The decline in catalytic activity was due to the sintering and carbon deposition under high reaction temperature. Although no decline was observed in the other three catalysts, sintering and carbon formation were inevitable during the long-term test and may influence the performance of catalysts in longer time catalysis. The coking resistance abilities of the samples were primarily compared through XRD patterns of spent catalysts (Fig. 8). New diffraction peaks were observed at 2q ¼ 26.4 , 42.2 ,44.4 and 54.5 which were indexed to graphite-2H phase (JCPDS 41-1487). The intensity of graphite diffraction peaks indicated that carbon deposition in Ni/MgO-SiO2 was obviously much weaker than the other catalysts, representing the highest coking resistance among the investigated catalysts. In the other three spent catalysts, the diffraction peaks of graphite-2H were more intense than that of Ni/MgO-SiO2, suggesting a more severe carbon deposition. The crystallite size of Ni nanoparticles calculated by Scherrer's equation was 13 nm in spent Ni/MgO-SiO2, 14 nm in spent Ni/SiO2, 19 nm in spent IM-Ni/MgO-SiO2 and 19 nm in spent IM-Ni/SiO2. The Ni particle sizes calculated by Scherrer's equation partly reflected

Fig. 7 e The catalytic stability in DRM of the reduce catalysts: (A) CH4 conversion and (B) CO2 conversion during 50 h under 700  C. Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

9

Fig. 8 e XRD patterns of the spent (A) Ni/MgO-SiO2, (B) Ni/SiO2, (C) IM-Ni/MgO-SiO2 and (D) IM-Ni/SiO2 catalysts after DRM under 700  C for 50 h.

Fig. 9 e TG, DTG and DSC curves of (A) Ni/MgO-SiO2, (B) Ni/SiO2, (C) IM-Ni/MgO-SiO2 and (D) IM-Ni/SiO2 catalysts after DRM under 700  C for 50 h. Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

10

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

Fig. 10 e TEM images of the spent (A, B) Ni/MgO-SiO2, (C, D) Ni/SiO2, (E, F) IM-Ni/MgO-SiO2 and (G, H) IM-Ni/SiO2 catalysts after DRM under 700  C for 50 h.

a superior metal sintering resistance of Ni/MgO-SiO2 (see Fig. 9). TGA was employed to quantify the amount of carbon species generated during the 50-h catalysis as shown in

Fig. 8. Weight changes occurred below 400  C were attributed to the vaporization of adsorbed water in loss and the oxidation of Ni metal particles in gain. The amount of deposited carbon could be estimated from the distinct mass loss

Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

occurred from the top of the oxidation peaks at about 400  Ce950  C. The carbon amounts were about 21.1 wt% for Ni/MgO-SiO2 and 64.8 wt% for Ni/SiO2. The superior coking resistance ability of Ni/MgO-SiO2 was attributed to the addition of MgO. The alkalinity of MgO effectively helps to adsorb CO2 species on the catalyst, suppressing the disproportionation reaction of CO. Meanwhile, the adsorbed CO2 reacted with newly formed carbon and helped to eliminate the deposited filamentous form carbon. The effect of MgO addition in coking resistance might also be inferred through comparing the carbon deposition amounts of IM-Ni/MgOSiO2 and IM-Ni/SiO2. Despite the same solid sphere structure, impregnated Ni particles and poor surface area, the carbon deposition amount was about 45.3 wt% in IM-Ni/MgO-SiO2 and 77.1 wt% in IM-Ni/SiO2. Although presenting similar composition, the coking resistance ability of Ni/MgO-SiO2 was obviously better than IM-Ni/MgO-SiO2. Metal particle size and the interaction between Ni nanoparticles and SiO2 support played significant roles in carbon deposition resistance during DRM [64,65]. According to previous reports, carbon formation occurred only when Ni particle size was larger than 9 nm, and larger metal particles contributed to the coke formation on Ni catalysts [13]. Strong metal-support interaction in the hydrothermal product Ni/MgO-SiO2, which was proved by the TPR results (Fig. 6) and discussed in Section Effect of synthesis route on catalytic activity, effectively inhibited the migration and growth of Ni particles, hence reduced carbon formation during the 50-h catalysis. This theory also could be used to explain the better performance of hydrothermal sample Ni/ SiO2 than the performance of impregnated sample IM-Ni/SiO2. Structure and morphology of the catalysts after 50-h catalysis were compared with fresh samples through SEM (Fig. S8) and TEM images (Fig. 10). Carbon nanotubes were observed in all the spent catalysts, consistent with the XRD results (Fig. 7). Both SEM and TEM images confirmed that no severe deformation or collapse occurred in Ni/MgO-SiO2 sample, suggesting a high sintering resistance and structural stability. Surface morphology and porous sphere structure maintained after the long-term catalysis (Fig. S8A). The surface area of Ni/MgO-SiO2 declined to 387.3 m2/g as a result of carbon formation (Table 2). Only slight particle growth of Ni particles (from 6.6 nm to 8.3 nm) was observed in Ni/MgO-SiO2 after 50 h (Fig. 10A and B). The particle stability could be attributed to the strong metal-support interaction. In contrast, destruction of the hollow hierarchical spheres was observed in Ni/SiO2 as the result of support sintering and carbon nanotubes deposition. SEM images (Fig. S8B) indicated that the outside of the spheres was partially covered by the carbon nanotubes, which led to the decline in surface area (102.4 m2/ g). The TEM images (Fig. 10C and D) show the collapse of part of the hollow spheres and the aggregation into bulky SiO2. Meanwhile, the Ni nanoparticles dropped out from the destroyed carrier and were shrouded within the carbon nanotubes. The average Ni particle size in Ni/SiO2 grew form 8.6 nme12.5 nm during the long-term catalysis (Table 1). However, some Ni nanoparticles were shrouded by the carbon nanotubes or covered by the bulky SiO2, which might lead to deactivation or even complete inactivation of these Ni particles.

11

The superiority in carrier stability of Ni/MgO-SiO2 can be attributed to the high coking resistance and formation of Mgphyllosilicate Mg3Si4O10(OH)2. On one hand, the high coking resistance suggested by TGA effectively prohibited structural damage by excessive carbon nanotubes and the detachment of Ni metal particles. On the other hand, the crystallized Mg3Si4O10(OH)2 was believed to play a role of the skeleton, helping to stabilize the amorphous SiO2 support at high temperatures and inhibit carrier sintering in Ni/MgO-SiO2. Bian et al. reported a similar structure deformation phenomenon in their Ni-Mg phyllosilicate nanotubes (PSNTS), which were not thermally stable and would decompose at 680e700  C. A SiO2 coating was employed to obtain a high thermal stability by the confinement effect from the mesoporous silica shell [14]. Considering the same chemical substances between Ni/MgOSiO2 and PSNTS, the hierarchical hollow sphere structure in our work has an advantage of higher thermal stability. The high thermal stability of the hollow hierarchical Ni/MgO-SiO2 can be obtained merely by a facile hydrothermal reaction, rather than a two-step process in PSNTS@silica structure. In addition, the hollow hierarchical structure yielded 30% higher surface area than the PSNTS@silica structure. In the impregnation samples, IM-Ni/MgO-SiO2 and IM-Ni/ SiO2, metal sintering and carbon deposition were also observed through SEM (Fig S8C-D) and TEM images (Fig. 10EeH). Ni nanoparticles shrouded within the carbon nanotubes were also observed in IM-Ni/MgO-SiO2 and IM-Ni/ SiO2. Ni particles in impregnation samples presented weak metal-support interaction as TPR results showed (Fig. 6), and they were more likely to detach from the SiO2 support. Moreover, especially severe carbon formation occurred in spent IM-Ni/SiO2 a, where SiO2 spheres were almost totally wrapped in the carbon nanotubes. The transport and contact of reaction gases on SiO2 support were severely hampered. Moreover, the adsorption and activation of reaction gases was difficult to continue as a result that Ni particles on SiO2 were covered by carbon nanotubes. The coverage of carbon nanotubes was considered to be chief culprit of the catalytic deactivation in Ni/SiO2 (Fig. 7).

Conclusions In conclusion, we have synthesized an MgO promoted Ni/ MgO-SiO2 catalyst through one-pot hydrothermal method. The high surface area of the hierarchical hollow sphere structure greatly promoted gas diffusion and increased the dispersity of Ni nanoparticles. Ni nanoparticles with small particle size and strong metal-support interaction were obtained through the exsolution from Ni-phyllosilicate during H2 reduction. The highly dispersed Ni nanoparticles provided sufficient reaction active sites and inhibit high carbon deposition in Ni/MgO-SiO2 catalyst, resulting in a superior catalytic activity and particle sintering resistance to the impregnated IM-Ni/SiO2 and IM-Ni/MgO-SiO2 samples. Moreover, the addition of Mg2þ in Ni/MgO-SiO2 during hydrothermal process promoted the SiO2 dissolution and phyllosilicates formation, leading to large surface area than the hollow hierarchical Ni/ SiO2. The uniformly dispersed MgO basic promoter increased CO2 adsorbing affinity, and the formation of Mg-phyllosilicate,

Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

12

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

providing channels for carrier activation of CO2 and improving the catalytic activity and coking resistance. In addition, the Mg-phyllosilicate showed a skeleton effect in Ni/MgO-SiO2, helped to stabilize the amorphous SiO2 support and to inhibit sintering of the support, leading to a good thermostability at high temperatures.

Acknowledgments This work was supported by the National Natural Science Foundation of China (No. 51672174), Natural Science Foundation of Shanghai Municipality (No. 17ZR1414900), Youth Program of National Natural Science Foundation of China (No. 51502170), the Material Genome Initiative Project Foundation of Science and Technology Commission of Shanghai Municipality (No. 16DZ2260602), and the support from the Advanced Energy Material and Technology Center of Shanghai Jiao Tong University.

Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.ijhydene.2018.05.010.

references

[1] Wang W, Wang SP, Ma XB, Gong JL. Recent advances in catalytic hydrogenation of carbon dioxide. Chem Soc Rev 2011;40:3703e27. [2] Song CS. Global challenges and strategies for control, conversion and utilization of CO2 for sustainable development involving energy, catalysis, adsorption and chemical processing. Catal Today 2006;115:2e32. [3] Bernard S, Horsfield B. Thermal maturation of gas shale systems. Annu Rev Earth Planet Sci 2014;42:635e51. [4] Melikoglu M. Shale gas: analysis of its role in the global energy market. Renew Sustain Energy Rev 2014;37:460e8. [5] Wang N, Yu X, Shen K, Chu W, Qian W. Synthesis, characterization and catalytic performance of MgO-coated Ni/SBA-15 catalysts for methane dry reforming to syngas and hydrogen. Int J Hydrogen Energy 2013;38:9718e31. [6] Liu D, Xian YQ, Wei NEC, Lau R, Borgna A, Yang Y. MCM-41 supported nickel-based bimetallic catalysts with superior stability during carbon dioxide reforming of methane: effect of strong metalesupport interaction. J Catal 2009;266:380e90. [7] Assabumrungrat S, Charoenseri S, Laosiripojana N, Kiatkittipong W, Praserthdam P. Effect of oxygen addition on catalytic performance of Ni/SiO$MgO toward carbon dioxide reforming of methane under periodic operation. Int J Hydrogen Energy 2009;34:6211e20. [8] Lovell EC, Fuller A, Scott J, Amal R. Enhancing Ni-SiO2 catalysts for the carbon dioxide reforming of methane: reduction-oxidation-reduction pre-treatment. Appl Catal, B 2016;199:155e65. [9] Tsoukalou A, Imtiaz Q, Kim SM, Abdala PM, Yoon S, Muller CR. Dry-reforming of methane over bimetallic Ni-M/ La2O3 (M ¼ Co, Fe): the effect of the rate of La2O2CO3 formation and phase stability on the catalytic activity and stability. J Catal 2016;343:208e14.

[10] Verykios XE. Catalytic dry reforming of natural gas for the production of chemicals and hydrogen. Int J Hydrogen Energy 2003;28:1045e63.  lvez ME. Ceria and [11] Liu H, Costa PD, Taief HBH, Benzina M, Ga zirconia modified natural clay based nickel catalysts for dry reforming of methane. Int J Hydrogen Energy 2017;37:23508e16. [12] Pakhare D, Spivey J. A review of dry (CO2) reforming of methane over noble metal catalysts. Chem Soc Rev 2014;43:7813. [13] Sibudjing K, Yasotha K, Jun N, Usman O, Ziwei L, Eng Toon S. Progress in synthesis of highly active and stable nickel-based catalysts for carbon dioxide reforming of methane. Chemsuschem 2015;8:3556e75. [14] Bian Z, Suryawinata IY, Kawi S. Highly carbon resistant multicore-shell catalyst derived from Ni-Mg phyllosilicate nanotubes@silica for dry reforming of methane. Appl Catal, B 2016;195:1e8. [15] Tsyganok AI, Tsunoda T, Hamakawa S, Suzuki K, Takehira K, Hayakawa T. Dry reforming of methane over catalysts derived from nickel-containing Mg-Al layered double hydroxides. J Catal 2003;213:191e203. [16] Cao Y, Lu M, Fang J, Shi L, Zhang D. Hexagonal boron nitride supported mesoSiO2-confined Ni catalysts for dry reforming of methane. Chem Commun 2017;53(54):7549e52. [17] Rouibah K, Barama A, Benrabaa R, Guerrero-Caballero J, Kane T, Vannier RN, et al. Dry reforming of methane on nickel-chrome, nickel-cobalt and nickel-manganese catalysts. Int J Hydrogen Energy 2017;42(50):29725e34. [18] Pino L, Italiano C, Vita A, Lagana M, Recupero V. Ce0.70La0.20Ni0.10O2-delta catalyst for methane dry reforming: influence of reduction temperature on the catalytic activity and stability. Appl Catal, B 2017;218:779e92. [19] AlSabban B, Falivene L, Kozlov SM, Aguilar-Tapia A, OuldChikh S, Hazemann JL, et al. In-operando elucidation of bimetallic CoNi nanoparticles during high-temperature CH4/ CO2 reaction. Appl Catal, B 2017;213:177e89. [20] Guo J, Lou H, Zhao H, Chai D, Zheng X. Dry reforming of methane over nickel catalysts supported on magnesium aluminate spinels. Appl Catal A 2004;273:75e82. [21] Yang W, Liu H, Li Y, Zhang J, Hao W, He D. Properties of yolkeshell structured Ni@SiO 2 nanocatalyst and its catalytic performance in carbon dioxide reforming of methane to syngas. Catal Today 2016;259:438e45. [22] Erdogan B, Arbag H, Yasyerli N. SBA-15 supported mesoporous Ni and Co catalysts with high coke resistance for dry reforming of methane. Int J Hydrogen Energy 2018;43:1396e405. [23] Khalesi Alieh, Arandiyan Hamid, Parvari Matin. Effects of lanthanum substitution by strontium and calcium in La-NiAl perovskite oxides in dry reforming of methane. Chin J Catal 2008;29:960e8. [24] Li X, Li D, Tian H, Zeng L, Zhao ZJ, Gong J. Dry reforming of methane over Ni/La2O3 nanorod catalysts with stabilized Ni nanoparticles. Appl Catal, B 2017;202:683e94. [25] Zhang Q, Wu T, Zhang P, Qi R, Huang R, Song X, et al. Facile synthesis of hollow hierarchical Ni/g-Al2O3 nanocomposites for methane dry reforming catalysis. RSC Adv 2014;4:51184e93. [26] Luna AEC, Iriarte ME. Carbon dioxide reforming of methane over a metal modified Ni-Al2O3 catalyst. Appl Catal A 2008;343:10e5. [27] Xu Z, Zhen M, Bi Y, Zhen K. Catalytic properties of Ni modified hexaaluminates LaNi y Al12yO19d for CO2 reforming of methane to synthesis gas. Appl Catal A 2000;198:267e73.

Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 3

[28] Wu T, Cai W, Zhang P, Song X, Gao L. Cu-Ni@SiO2 alloy nanocomposites for methane dry reforming catalysis. RSC Adv 2013;3:23976e9. [29] Li Z, Kathiraser Y, Ashok J, Oemar U, Kawi S. Simultaneous tuning porosity and basicity of Nickel@Nickelemagnesium phyllosilicate coreeshell catalysts for CO2 reforming of CH4. Langmuir 2014;30:14694e705. [30] Hu YH. Solid-solution catalysts for CO2 reforming of methane *. Catal Today 2009;148:206e11. [31] Jafarbegloo M, Tarlani A, Mesbah AW, Muzart J, Sahebdelfar S. NiOemgo solid solution prepared by solegel method as precursor for Ni/MgO methane dry reforming catalyst: effect of calcination temperature on catalytic performance. Catal Lett 2016;146:238e48. [32] Cj Liu, Ye J, Jiang J, Pan Y. Progresses in the preparation of coke resistant Ni-Based catalyst for steam and CO2 reforming of methane. ChemCatChem 2011;3:529e41. [33] Xu L, Song H, Chou L. One-pot synthesis of ordered mesoporous NiOeCaOeAl2O3 composite oxides for catalyzing CO2 reforming of CH4. ACS Catal 2012;2:1331e42. [34] Zubenko D, Singh S, Rosen BA. Exsolution of Re-alloy catalysts with enhanced stability for methane dry reforming. Appl Catal, B 2017;209:711e9. [35] Osaki T, Mori T. Role of potassium in carbon-free CO2 reforming of methane on K-Promoted Ni/Al2O3 catalysts. J Catal 2001;204:89e97. [36] Frusteri F, Spadaro L, Arena F, Chuvilin A. TEM evidence for factors affecting the genesis of carbon species on bare and Kpromoted Ni/MgO catalysts during the dry reforming of methane. Carbon 2002;40:1063e70. [37] Dias JAC, Assaf JM. Influence of calcium content in Ni/CaO/gAl2O3 catalysts for CO2 -reforming of methane. Catal Today 2003;85:59e68. [38] Hou Z, Yokota O, Tanaka T, Yashima T. Characterization of Ca-promoted Ni/a-Al2O3 catalyst for CH4 reforming with CO2. Appl Catal A 2003;253:381e7. [39] Koo KY, Roh HS, Yu TS, Dong JS, Wang LY, Park SB. Coke study on MgO-promoted Ni/Al2O3 catalyst in combined H2O and CO2 reforming of methane for gas to liquid (GTL) process. Appl Catal A 2008;340:183e90. [40] Miyamoto M, Hamajima A, Oumi Y, Uemiya S. Effect of basicity of metal doped ZrO2 supports on hydrogen production reactions. Int J Hydrogen Energy 2018;43:730e8. [41] Sutthiumporn K, Kawi S. Promotional effect of alkaline earth over NieLa2O3 catalyst for CO2 reforming of CH4: role of surface oxygen species on H2 production and carbon suppression. Int J Hydrogen Energy 2011;36:14435e46. [42] Zhao X, Li H, Zhang J, Shi L, Zhang D. Design and synthesis of NiCe@m-SiO2 yolk-shell framework catalysts with improved coke- and sintering-resistance in dry reforming of methane. Int J Hydrogen Energy 2016;41:2447e56. [43] Zhao X, Lu M, Li H, Fang J, Shi L, Zhang D. In situ preparation of Ni nanoparticles in ceriumemodified silica aerogels for coking- and sintering-resistant dry reforming of methane. New J Chem 2017;41(12):4869e78. [44] Pan Y, Kuai P, Liu Y, Ge Q, Liu C. Promotion effects of Ga2O3 on CO2 adsorption and conversion over a SiO2-supported Ni catalyst. Energy Environ Sci 2010;3:1322e5. [45] Xu L, Song H, Chou L. Carbon dioxide reforming of methane over ordered mesoporous NiOeMgOeAl 2 O 3 composite oxides. Appl Catal, B 2011;108:177e90. [46] Shiraz MHA, Rezaei M, Meshkani F. The effect of promoters on the CO2 reforming activity and coke formation of nanocrystalline Ni/Al2O3 catalysts prepared by microemulsion method. Kor J Chem Eng 2016;33:1e8. [47] Narasimharao K, Ali TT, Bawaked S, Basahel S. Effect of Si precursor on structural and catalytic properties of nanosize magnesium silicates. Appl Catal A 2014;488:208e18.

13

[48] Wang Y, Li B, Wang Q, Shou Z. Development of a cataluminescence sensor for detecting benzene based on magnesium silicate hollow spheres. Luminescence 2015;30:619. [49] Shalvoy R, Reucroft P, Davis B. Characterization of coprecipitated nickel on silica methanation catalysts by Xray photoelectron spectroscopy. J Catal 1979;56:336e48. [50] Wu T, Zhang Q, Cai W, Zhang P, Song X, Sun Z, et al. Phyllosilicate evolved hierarchical Ni- and CueNi/SiO 2 nanocomposites for methane dry reforming catalysis. Appl Catal A 2015;503:94e102. [51] Seyama H, Soma M. X-ray photoelectron spectroscopic study of montmorillonite containing exchangeable divalent cations. J Chem Soc, Faraday Trans 1: Phys Chem Condens Phase 1984;80(1):237e48. [52] Ghods B, Rezaei M, Meshkani F. Ni catalysts supported on mesoporous nanocrystalline magnesium silicate in dry and steam reforming reactions. Chem Eng Technol 2017;40(4):760e8. [53] Kathiraser Y, Thitsartarn W, Sutthiumporn K, Kawi S. Inverse NiAl2O4 on LaAlO3eAl2O3: unique catalytic structure for stable CO2 reforming of methane. J Phys Chem C 2013;117:8120e30. [54] Guo JJ, Lou H, Zhao H, Chai DF, Zheng XM. Dry reforming of methane over nickel catalysts supported on magnesium aluminate spinels. Appl Catal A 2004;273:75e82. [55] Jin L, Ma B, Zhao S, He X, Li Y, Hu H, et al. Ni/MgO-Al2O3 catalyst derived from modified [Ni, Mg, Al]-LDH with NaOH for CO2 reforming of methane. Int J Hydrogen Energy 2018;43(5):2689e98. [56] Fan MS, Abdullah AZ, Bhatia S. Catalytic technology for carbon dioxide reforming of methane to synthesis gas. ChemCatChem 2009;1:192e208. [57] Chen YG, Yamazaki O, Tomishige K, Fujimoto K. Noble metal promoted Ni0.03Mg0.97O solid solution catalysts for the reforming of CH4 with CO2. Catal Lett 1996;39:91e5. [58] Bitter J, Seshan K, Lercher J. Mono and bifunctional pathways of CO2/CH4 reforming over Pt and Rh based catalysts. J Catal 1998;176:93e101. [59] Ferreira-Aparicio P, Rodrıguez-Ramos I, Anderson J, Guerrero-Ruiz A. Mechanistic aspects of the dry reforming of methane over ruthenium catalysts. Appl Catal A 2000;202:183e96. [60] Wei QH, Gao XH, Liu GG, Yang RQ, Zhang HB, Yang GH, et al. Facile one-step synthesis of mesoporous Ni-Mg-Al catalyst for syngas production using coupled methane reforming process. Fuel 2018;211:1e10. [61] Tang S, Ji L, Lin J, Zeng HC, Tan KL, Li K. CO2 reforming of methane to synthesis gas over solegel-made Ni/g-Al2O3 catalysts from organometallic precursors. J Catal 2000;194:424e30. [62] Ni J, Chen L, Lin J, Kawi S. Carbon deposition on borated alumina supported nano-sized Ni catalysts for dry reforming of CH4. Nano Energy 2012;1:674e86. [63] Taherian Z, Yousefpour M, Tajally M, Khoshandam B. Catalytic performance of Samaria-promoted Ni and Co/SBA15 catalysts for dry reforming of methane. Int J Hydrogen Energy 2017:42. [64] Chen D, Lødeng R, Anundska˚s A, Olsvik O, Holmen A. Deactivation during carbon dioxide reforming of methane over Ni catalyst: microkinetic analysis. Chem Eng Sci 2001;56:1371e9. [65] Du X, Zhang D, Shi L, Gao R, Zhang J. Morphology dependence of catalytic properties of Ni/CeO2 nanostructures for carbon dioxide reforming of methane. J Phys Chem C 2012;116:10009e16.

Please cite this article in press as: Zhang Q, et al., Hollow hierarchical Ni/MgO-SiO2 catalyst with high activity, thermal stability and coking resistance for catalytic dry reforming of methane, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/ j.ijhydene.2018.05.010