12CaO∙7Al2O3 catalysts: Effect of encaged anions on catalytic mechanism

Journal Pre-proof CO2 Methanation over Ru/12CaOnull7Al2 O3 Catalysts: Effect of Encaged Anions on Catalytic Mechanism Hongjie Wu, Mingwei Yuan, Jia Huang, Xiaozhong Li, Yan Wang, Jinjun Li, Zhixiong You

PII:

S0926-860X(20)30067-3

DOI:

https://doi.org/10.1016/j.apcata.2020.117474

Reference:

APCATA 117474

To appear in:

Applied Catalysis A, General

Received Date:

4 December 2019

Revised Date:

10 February 2020

Accepted Date:

21 February 2020

Please cite this article as: Wu H, Yuan M, Huang J, Li X, Wang Y, Li J, You Z, CO2 Methanation over Ru/12CaOx2219;7Al2 O3 Catalysts: Effect of Encaged Anions on Catalytic Mechanism, Applied Catalysis A, General (2020), doi: https://doi.org/10.1016/j.apcata.2020.117474

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Published by Elsevier.

CO2 Methanation over Ru/12CaO∙7Al2O3 Catalysts: Effect of Encaged Anions on Catalytic Mechanism Hongjie Wu,a,b Mingwei Yuan, a,b Jia Huang, a,b Xiaozhong Li, a,b Yan Wang, a,bJinjun Li, a Zhixiong You*,a,b a

oo f

School of Resources and Environmental Sciences, Wuhan University, Wuhan 430079, P.R.

China. E-mail: [email protected]; Fax:+86 27-68778893; Tel: +86 27-68778936

International Cooperation Base for Sustainable Utilization of Resources and Energy in Hubei

pr

b

e-

Province, Wuhan 430072, P.R. China

Jo

ur

na l

Pr

Graphical abstract

Highlights: 

OH- or e- encaged in mayenites promotes the activation of CO2 on the supports.

1



The catalytic activity of Ru/C12A7:OH- is drastically higher than that of Ru/C12A7:e- and Ru/-Al2O3.

*CO adsorbed on Ru is the vital intermediate for methane formation and carbon deposition.



Electrons in mayenite electrides are transferred to Ru metal, thus accelerating the

oo f



pr

dissociation of *CO into *C (carbon deposition).

e-

Abstract:

Pr

Mayenites (C12A7) encaged with OH- or e- have been synthesized and applied as supports for ruthenium catalyst in CO2 methanation. The catalytic activity of Ru/C12A7:OH- (TOF at 225 oC

na l

= 6.04  10-2 s-1, Ea = 70.9 kJ mol-1) is drastically higher than that of Ru/C12A7:e- (TOF at 225 oC = 2.41  10-2 s-1, Ea = 94.4 kJ mol-1) and Ru/-Al2O3 (TOF at 225 oC = 0.23  10-2 s-1, Ea = 105.8

ur

kJ mol-1). It has been proposed that the OH- or e- encaged in mayenite promotes the formation of carbonates and bicarbonates from CO2 on the supports. It is supposed that at the interface between

Jo

Ru and mayenite support the surface carbonates and bicarbonates are hydrogenated into formates, which are further dissociated into *CO adsorbed on Ru. The formed *CO is the intermediate for methane formation and carbon deposition. The different anions (OH- or e-) encaged in the framework of mayenites have been found to influence drastically the ionization potential of Ru. 2

Electrons in mayenite electrides are transferred to Ru metal, thereby enhancing the back donation from Ru to the antibonding orbitals of *CO intermediate. As a consequence, the dissociation of *

CO into *C and subsequent into deposition of carbons are accelerated, which is responsible for

the inferior catalytic performance of Ru/C12A7:e-. Based on the experimental results, we propose an effective technique for tailoring the catalytic performance of metallic sites supported on

oo f

mayenites via changing encaged anions.

e-

pr

Keywords: CO2 methanation, ruthenium catalyst, mayenite, reaction mechanism

Pr

1. Introduction

CO2 concentration in the atmosphere is reported to rise from pre-industrial age’s ~280 ppm

na l

to hitherto ~400 ppm [1], which leads to several serious consequences: the change of global climate, the acidification of oceans, and the rising of sea levels [2]. Burning fossil fuels is the

ur

main anthropogenic emission of CO2. In the near future, fossil fuels will still be the dominant primary energy that we can rely on; although the emission of CO2 would be mitigated

Jo

significantly by switching to renewable energies [3]. Therefore, efficient and economic conversion of CO2 into value-added chemicals or fuels would contribute concurrently to reducing CO2 emission and easing the dependence of human being on fossil fuels. However, renewable energy resources such as wind, solar, tide are weather-dependent, which makes the integration of renewable energies into the existing power grid difficult. As a solution to this 3

problem, a power to gas (P2G) technology is proposed. The P2G technology involves: (i) conversion of electric power generated from intermittent and fluctuating renewable energies to hydrogen via electrolysis and (ii) production of substitute natural gas (SNG) by hydrogenation of CO2 to methane [3-5]. However, CO2 is emitted in low concentration and complex composition from combustion plants or power generation plants [6,7]. Thus, CO2 in flue gas should be

oo f

purified and concentrated before it is catalytically converted into SNG [6]. Recently a

combination of CO2 capture and catalytic hydrogenation on solid materials has been proposed

pr

[7]. Here we propose to use mayenite encaged with OH- or e- as support for Ru catalysts for CO2

e-

methanation with an intent to enhance the adsorption and activation of CO2 on the supports. CO2 methanation proceeding according to equation (1) is also called Sabatier reaction [8].

Pr

At atmospheric pressure, CO and CH4 are the only two C-containing products over most metal catalysts [8, 9].

na l

CO2 + 4H2 ↔ CH4 +2H2O (Ho = -165 kJ mol-1)

(1)

This reaction is exothermic and thermodynamically favored at low temperatures where

ur

kinetic rates are low. Thus, a catalyst is required to activate CO2 and H2 to obtain an acceptable

Jo

reaction rate and selectivity at temperatures lower than 350 oC. At temperatures above 350 oC, methane reforming and reverse water gas shift (RWGS) reaction, which are both endothermic, become thermodynamically favored. In the past few decades, a great number of catalysts have been investigated for CO2 methanation. Active metals, for example Ni [10-14], Ru [15, 16], Rh [17, 18], Pd [19-21], Fe 4

[22, 23], or Co [22], have been identified. Supports, such as Al2O3 [19], SiO2 [24, 25], ZrO2 [11], CeO2[12, 26], La2O3 [27], MgO [28],TiO2 [15, 18], and zeolites [13], have been suggested to carry the active metals for CO2 methanation. Among these active metals, Ru-based catalysts seem to be a good candidate for activating CO2 in hydrogen atmosphere due to its high specific activity and resistance to carbon deposition.

oo f

[Ca24Al28O64]4+(e-)4 (C12A7:e- for brevity) [29, 30], is a water-vapor and air tolerant

electride , in which electrons serve as anions. In 2003, C12A7:e- was first synthesized from

pr

mayenite (12CaO·7Al2O3). The chemical formula of a unit cell of mayenite is expressed as

e-

[Ca24Al28O64]4+(O2-)2 (C12A7 for brevity) [29]. The crystal of C12A7 is composed of a framework of interconnected cages, each cage in a size of ~0.4 nm bears positive charge of +1/3,

Pr

which is compensated by anions, such as OH-, Cl-, H-, Au-, O-, O2-, S2-, O22-, and e-, which are moving around the cages [31-34]. Masaaki Kitano et al. [35] found that C12A7:e- is a favorable

na l

electron donor and reversible hydrogen store in a Ru-loaded catalyst for ammonia synthesis. Yoshitake Toda et al. [36] reported that the surface of C12A7:e- can activate and split CO2 at

ur

room temperature due to a high concentration of encaged electrons in the near-surface region. However, the CO(ads) molecules, originated from CO2 decomposition on the surface of

Jo

C12A7:e-, mainly desorbed at T > 800 ℃. With a intent to apply the outstanding catalytic performance of C12A7:e- in CO2 activation, we try to support Ru nanoparticles on C12A7 (with OH- or e- in the cages of mayenite) and to provide a high energy reactant, hydrogen, to accelerate the conversion of CO2 to methane instead of to CO. 5

So far, there is still no clear consensus on the actual reaction path of CO2 methanation despite numerous efforts have been devoted to this field [16, 37]. The answer to the question whether CO(ads) is the intermediate or not of CO2 methanation may vary depending on the nature of catalysts, or even on the partial pressure of CO2 and H2. Here, we reveal that anions (OH- or e-) encaged in C12A7 plays a key role in the reaction mechanism of CO2 methanation

oo f

over Ru nanoparticles via operando FTIR, CO2-TPD, and temperature programmed surface

pr

reaction studies carried out at transient or steady states.

e-

2. Experimental Section 2.1 Support synthesis

Pr

[Ca24Al28O64]4+(OH-)4 (C12A7:OH- for brevity) was synthesized by calcination of hydrothermally treated mixture of hydroxide of Ca and Al [38]. Briefly, a stoichiometric (Ca:Al

na l

= 12:14) mixture of Ca(OH)2 (analytical grade, Sinopharm Chemical Reagent, China) and Al(OH)3 (analytical grade, Sinopharm Chemical Reagent, China) was dispersed in pure water

ur

and ground with an agate mortar and pestle for 1 h. The ground mixture was transferred into a Teflon-lined autoclave (500 mL) and treated at 150 ℃ for 6 h with stirring at 360 rpm. The

Jo

hydrothermally treated product was then dried in a rotary evaporator (10 KPa, 50 ℃). The dried sample was calcined in air at 400 – 900 ℃ (10 ℃ min-1) for 5 h. The obtained C12A7:OHsupport is denoted with calcination temperature in front of its abbreviation as 400C12A7:OH-, 500C12A7:OH-, 600C12A7:OH-, 700C12A7:OH-, 800C12A7:OH- or 900C12A7:OH-, 6

respectively. C12A7:e- was achieved by reducing the resultant C12A7:OH- powder with CaH2 (95 %, Sinopharm Chemical Reagent, China) in an Ar atmosphere . The 600C12A7:OH- powder (calcined at 600 ℃ for 5 h) was first dehydroxylated at 900 ℃ for 15 h in an Ar (99.999%) stream of 5 mL min-1. The dehydroxylated powder (2.0 g) was ground with excess amount of CaH2

oo f

(95 %, 0.640 g) in Ar atmosphere in a glovebox. The mixture was heated at 700 – 1000 ℃ for 15 h in an Ar (99.999 %) stream of 5 mL min-1. The reduced sample was labeled with the reduction

pr

temperature in front of its name as 700C12A7:e-, 800C12A7:e-, 900C12A7:e- and 1000C12A7:e-,

e-

respectively.

-Al2O3 support was obtained by calcining a commercial alumina at 900 ℃ for 5 h in static

Pr

air.

na l

2.2 Catalyst preparation

Ruthenium was supported by wet impregnation method. The support (C12A7:OH-, -Al2O3,

ur

or C12A7:e-) was dispersed in an anhydrous tetrahydrofuran (THF, Sinopharm Chemical Reagent, China) solution of Ru3(CO)12 (99.0 %, Aldrich, USA) and stirred at room temperature

Jo

for 12 h. Then the solvent was removed with a rotary evaporator (40 ℃, 10 KPa). Subsequently the dried powder was sealed in a quartz tube and heated in an Ar (99.999 %) stream of 5 mL min1

following a temperature program as: 0.5 ℃ min-1 up to 60 ℃, hold for 2 h; 0.5 ℃ min-1 up to

120 ℃, hold for 1 h; 0.5 ℃ min-1 up to 200 ℃, hold for 2 h; down to ambient temperature. The 7

catalysts obtained had a nominal Ru loading of 2 wt%. In addition, a nominal Ru loading of 1, 3 and 4 wt% on 500C12A7:OH- was prepared to study the size effect of Ru particles on the catalytic activity for CO2 methanation.

2.3 Characterization

oo f

The N2 adsorption-desorption isotherms of the samples were recorded at -196 ℃ on an automatic specific surface area/pore size distribution apparatus (BELSORP-mini Ⅱ, BEL,

pr

Japan). Prior to tests, the samples were evacuated (10-2 KPa) at 200 ℃ for 2 h. The specific

e-

surface areas were calculated from the linear part of the BET plot where P/P0 ratios are 0.05 – 0.25. The pore volume and mean pore diameter were calculated using BJH model. The

Pr

crystalline structure of the samples was characterized on a powder X-ray diffractometer (X’Pert Pro, PANalytical, England) equipped with a Cu-K radiation source ( = 1.5405 Å). All

na l

diffraction patterns were scanned in a 2 range of 10 – 90o at a rate of 2o min-1. The morphology of the supports was observed on a scanning electron microscopy (Zeiss SIGMA, Carl Zeiss AG,

ur

Germany) at an acceleration voltage of 20 kV. The images of Ru nanocrystallites on the supports

Jo

were recorded on a transmission electron microscopy (JEM-2100, JEOL, Japan) at an acceleration voltage of 200 kV. To analyze the real loading of Ru, appropriate amount of the synthesized catalyst was dissolved in 4 mL 18 wt % hydrochloric acid in an autoclave at 150 oC for 4 h. The hydrothermally dissolved Ru solution was appropriately diluted with pure water and measured by an inductively coupled plasma-atomic emission spectrum (ICP-AES, IRIS Intrepid 8

II XSP, Thermo Fischer Scientific Inc., USA). The UV-Visible optical absorption spectra of samples were obtained by Kubelka-Munk transformation of diffuse reflectance spectra on a UVVis spectrophotometer (UV-2550, Shimadzu, Japan). X-ray photoelectron spectra were recorded using a spectrometer with a monochromatic Al-K source (1486.6 eV) (XPS, NEXSA, Thermo Fisher Scientific Inc., USA). The base pressure of the analytical chamber was ~10-7 Pa.

oo f

Sputtering was performed using an Ar ion beam (6.4 mA, 3.5 kV) for 50 s, and the XPS spectra were recorded at sputtering time of 0, 20, and 50 s. Ru dispersion was determined by a CO-pulse

pr

titration method performed at 50 oC on an semi-automatic gas-adsorption instrument (DAS-7000,

e-

Huasi Instrument Co.,Ltd, China). Prior to CO-pulse titrations, the samples (100.0 mg) were pretreated in 50 mL min-1 He (99.999 %) at 300 ℃ for 30 min, and then reduced in 50 mL min-1

Pr

H2 (99.999 %) at 300 ℃ for 4 h. Hydrogen atoms adsorbed on the reduced catalysts were purged with 50 mL min-1 He (99.999 %) at 300 ℃ for 30 min. After cooling to 50 oC under He, CO-pulse

na l

measurements were performed by repeatedly injecting 1.0 mL of 5 vol% CO in He (one pulse) at 50 oC until a constant TCD response was achieved. A stoichiometry of CO : Ru = 1 : 1 was

Jo

ur

assumed to calculate the metal dispersion.

2.4 CO2 methanation activity measurements The catalytic activities were tested at atmospheric pressure using a fixed bed plug-flow

system (Figure S1). Generally, 50.0 mg of catalyst (pellets in a size of 0.22 – 0.45 mm) was diluted with 500.0 mg quartz sand (in a size of 0.22 – 0.45 mm) and loaded into a quartz reactor 9

(i.d. = 8 mm) and sandwiched between two quartz wool blocks. A quartz thermowell (O.D. = 3 mm, I.D. = 2 mm) that penetrated through the catalyst bed was fixed in the center of the quartz reactor as shown in Figure S1. The reaction temperature was measured at the middle point of the catalyst bed via a K-type thermal couple (O.D. = 1.5 mm), which was inserted in the quartz thermowell. The reactant gases were fed via mass flow controllers (Sevenstar D07-19B/ZM,

oo f

Beijing Sevenstar Electronics Ltd., China). Prior to catalytic measurements, the catalysts were reduced at 300 ℃ for 5 h in a mixture of H2 (99.999 %) and Ar (99.999 %) (H2 : Ar = 2 : 1, 60

pr

mL min-1). After reduction, the reactor was cooled down to 150 ℃ in 20 mL min-1 of Ar flow.

e-

During activity tests, the feed was changed to 40 mL min-1 of H2 (99.999 %) + 10 mL min-1 of CO2 (99.99 %) + 10 mL min-1 of Ar (99.999 %, internal standard). The reaction temperature was

Pr

varied from 150 to 500 oC at 0.1 MPa. The effluent gas was passed through a cold-trap (ethylene glycol + liquid nitrogen at around -10 oC) to condense the produced water vapor. The

na l

composition of the products was analyzed by a gas chromatograph (GC) (GC-2020, Trustworthy, China) equipped with a TDX-01 packed column (2 mm (i.d.) × 2 m) and a thermal conductivity

ur

detector (TCD). The activity data reported herein are an average of three consecutive

Jo

measurements over 60 min at each reaction condition. No activity of the quartz wool and quartz sand was confirmed by blank test. The conversion of CO2 and yield to CH4 were calculated according to the following equations: 𝐶𝑂2 conversion (%) = 𝐶𝐻4 yield (%) =

𝐶𝑂2 𝑐𝑜𝑛𝑣𝑒𝑟𝑡𝑒𝑑 𝐶𝑂2 𝑓𝑒𝑑

𝐶𝐻4 𝑝𝑟𝑜𝑑𝑢𝑐𝑒𝑑 𝐶𝑂2 𝑐𝑜𝑛𝑣𝑒𝑟𝑡𝑒𝑑

× 100%

× 100%

(2) (3)

10

2.5 Temperature-programmed surface reaction (TPSR) Temperature-programmed surface reaction of CO2 or CO hydrogenation was conducted on a fix bed reactor coupled with a quadrupole rod mass spectrometer (HPR20, HIDEN Analytical Ltd., England) at atmospheric pressure. The catalyst (100.0 mg, 0.22 – 0.45 mm pellets) was

oo f

loaded into a fixed bed quartz reactor. Before performing the tests, the catalysts were reduced at 300 ℃ for 5 h in 60 mL min-1 of 2H2 (99.999 %) + 1Ar (99.999 %). After reduction, the reactor

pr

was cooled down to 100 ℃ in 20 mL min-1 Ar. During TPR tests, the catalysts were heated from

e-

100 to 700 oC at a rate of 5 oC min-1 in a flow of 60 mL min-1 of a mixture of H2, CO2 (CO) and Ar (internal standard) with a molar ratio of H2 : CO2 (CO) : Ar = 4 : 1 : 1. The product gas was

Pr

passed through a cold-trap (around -10 oC) to remove water vapor and then led into the mass

na l

spectrometer through a controlled leak valve.

2.6 Temperature-programmed desorption (TPD) of CO2

ur

Temperature-programmed desorption of CO2 (CO2-TPD) was conducted on a fix bed reactor

Jo

coupled with a quadrupole rod mass spectrometer (HPR20, HIDEN Analytical Ltd., England) at atmospheric pressure. The catalyst (100.0 mg, 0.22 – 0.45 mm pellets) was loaded into a fixed bed quartz reactor. Prior to CO2-TPD measurements, the catalysts were first pretreated in 20 mL min-1 Ar (99.999 %) at 300 ℃ for 30 min, and then reduced in 40 mL min-1 H2 (99.999 %) and 20 mL min-1 Ar (99.999 %) at 300 ℃ for 4 h. The H atoms adsorbed on the reduced catalysts were 11

removed by purging with 20 mL min-1 Ar (99.999 %) at 300 ℃ for 30 min. Then the reactor was cooled down to 50 ℃ in 20 mL min-1 Ar (99.999 %). The adsorption of CO2 was carried out in 10 mL min-1 CO2 (99.99 %) at 50 oC for 1 h. The saturated samples were flushed with 10 mL min-1 Ar (99.999 %) for 3 h. The samples were heated from 50 to 900 oC at 5 oC min-1 in 10 mL min-1

oo f

Ar (99.999 %).

2.7 In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS)

pr

investigation

e-

In situ DRIFTS investigations were conducted using a high temperature reaction chamber (HVC, Harrick Scientific Products Inc., USA) coupled with a Fourier transform infrared (FTIR)

Pr

spectrometer (Nicolet IS50, Thermo Fisher, USA) equipped with a mercury cadmium telluride detector. Samples (ca 0.1 g) were placed without packing or dilution on a screen with pore size

na l

of 0.104 mm, through which reactant gases could pass, in the sample cup with controlled temperature in the HVC chamber. The reactant gases were fed by mass flow controllers

ur

(Sevenstar D07-19B/ZM, Beijing Sevenstar Electronics Ltd., China). Prior to the tests, the

Jo

samples were reduced at 300 ℃ for 5 h in 50 mL min-1 of 4H2 + 1Ar. For steady state experiments, the samples were cooled down in 10 mL min-1 of Ar to 50 oC. After collecting the background spectrum, the samples were heated from 50 to 700 ℃ in 60 mL min-1 of 4H2+1CO2 (CO) +1Ar. The in situ DRIFT spectrum at each temperature was recorded after 10 min stabilization at desired temperature by collecting 32 scans at a resolution of 4 cm-1. In the cases 12

of transient state experiments, the sample was heated to the desired reaction temperature in 10 mL min-1 Ar flow after reduction. After the background spectrum was collected, we varied the reactant gas intently and recorded the transient DRIFT IR spectra via collecting 6 scans at a resolution of 4 cm-1 in each 10 s.

oo f

3. Results and discussion 3.1 Structure of synthesized mayenite supports

pr

The as-prepared C12A7:OH- supports are mainly composed of mayenite phase

e-

(Ca12Al14O33) (Figures S2), and their crystallinity increases with calcination temperature, in good agreement with the gradual decreasing from 53 (Figure S3) to 17 m2 g-1 observed in the results of

Pr

BET specific surface area of the samples (Table 1). CaO phase is also identified in the C12A7:OH- supports when being calcined at temperatures higher than 600 oC (Figure S2). The

na l

morphology of the C12A7:OH- supports appears in polyhedrons (mayenite phase) and flakes (CaO) in irregular shape as observed with SEM (Figure S4). The polyhedrons gradually grow

ur

from around 1 to 2 m when raising calcination temperature from 400 to 900 oC (Figure S4). To

Jo

verify the encaged anions in the synthesized C12A7:OH- samples, we recorded their UV-Visible diffraction reflectance spectra as shown in Figure 1a. The C12A7:OH- supports all display an optical absorption edge at about 5.5 eV, which reveals that OH- groups occupies dominantly the cages in the framework of hydrothermally prepared C12A7:OH- samples [34]. No obvious difference among the samples calcined at different temperature is observed in their UV-Visible 13

absorption band, indicating that calcination temperature does not influence significantly the species and amount of encaged anions in C12A7:OH-. We chose the 600 oC calcined C12A7:OH- sample as starting material for further synthesis of electrides because we did not identify obvious CaO phase in its XRD pattern (Figure S2). After being subjected to dehydroxylation and reduction with CaH2, the C12A7:e- supports

oo f

display a similar crystalline structure (Figure S5) with that of the 600C12A7:OH- sample.

However, the diffraction intensities become stronger as the reduction temperature is raised,

pr

which is in accordance with the obvious growth of the electride particles (Figure S6) and the

e-

further shrinkage in their specific surface area and porosity (Table 1). To determining the electron density in the C12A7:e- supports, we recorded their UV-Visible diffuse reflectance

Pr

spectra as shown in Figure 1b. The band with its edge at 5.5 eV disappears and a broad absorption band appears at around 2.8 eV, which proves that the encaged OH- groups are

na l

completely removed and electrons are injected into the cages of C12A7.[29] Based on the report by Matsuishi et al.[39], the top (Esp) of the band located at around 2.8 eV shifts with the electron

ur

density (Ne) of C12A7:e- following the equation of Ne = [-(Esp - Esp0)/0.199]0.782 (where Esp0 =

Jo

2.83 eV at Ne ≈ 1×1018 cm-3). We calculated the electron densities of our as synthesized C12A7:esupports, which are listed in Figure 1b. The electron density decreases from 0.92 × 1021 cm-3 to 0.58 × 1021 cm-3 as raising the reduction temperature from 700 to 900 ℃, and reaches the maximum value of 1.12 × 1021 cm-3 when the sample was reduced at 1000 ℃. Besides the 2.8 eV band, another band with absorption edge at 4.2 eV can be found in the electride supports reduced 14

at temperatures lower than 1000 oC in Figure 1b. This absorption band is ascribed to the O2anions in the cages[34]. These results indicate that two types of anions, i.e. e- and O2-, exist in the cages of the C12A7:e- supports reduced at temperatures lower than 1000 oC, which also explains their lower electron densities.

oo f

3.2 Catalyst characterization

To avoid the reaction of electrides with water, we chose anhydrous THF as solvent to

pr

dissolve Ru3(CO)12 precursor in wet impregnation process. During the decomposition of

e-

Ru3(CO)12 into Ru, a loss of Ru metal was observed. The loading of Ru measured by ICP is shown in Table 2. The size of the Ru particles on the C12A7:OH- supports dependents

Pr

significantly on the Ru loading (Figure S7). Ru particles are well dispersed in a size of around 12 nm for all the 2 wt% Ru catalysts (Figure S7). Ru crystallites of the 1 wt% Ru/500C12A7:OH-

na l

catalyst (Figure S7 g) are < 1 nm, in agreement with its high dispersion (99.4%) measured by CO-pulse titration method (Table 2). Raising Ru loading to 3.54 wt% results in enlarged Ru

ur

nanoparticles in size of 2-5 nm (Figure S7 h). We measured the size of 500 Ru particles on each

Jo

catalyst and calculated the average size of Ru as shown in Table 2. Based on the average size, the dispersion of Ru was calculated according to the method reported by Borodziński [40]. Due to the shrinkage in the specific surface area of the C12A7:e- observed (Table 1) during dehydroxylation and reduction, the Ru particles on the electride grows obviously to 2-4 nm (Figure S8 and Table 2). The corresponding Ru dispersion of Ru/C12A7:e- is around half of that 15

of Ru/C12A7:OH- having similar Ru loading. The dispersion results calculated from TEM images are comparable to the CO/Ru values measured with CO-pulse titration method (Table 2).

3.3 Catalytic performance of Ru/C12A7:OH- and Ru/C12A7:eBefore activity tests, the catalysts were reduced in a mixture of 40 mL min-1 H2 and 20 mL

oo f

min-1 Ar at 300 oC for 5 h, and then treated in reactant stream of 40 mL min-1 H2 + 10 mL min-1 CO2 + 10 mL min-1 Ar at 500 oC for 5 h to ensure reaching a quasi-stable state, which was

pr

verified by no obvious difference in the activity data measured by increasing and decreasing

e-

temperature scan as shown in Figure S9. Figure 2 compares the catalytic activity of CO2 methanation over Ru catalysts supported on C12A7:OH-, C12A7:e-, or -Al2O3. All the tested

Pr

catalysts demonstrate a selectivity to CH4 higher than 99 % at temperatures lower than 425 ℃ (Figures 2c and d), and the selectivity to CO increases gradually to around 5 % when reaction

na l

temperature is raised to 500 oC, at which methane reforming and RWGS reactions become thermodynamically favorable. No other C-containing products such as oxygenated C1 or C2+

ur

hydrocarbons are detected in both the condensates and product gases. Figure 2 exhibits the

Jo

catalytic performance of typical Ru catalysts supported on C12A7:OH-, -Al2O3 and C12A7:e-, the curves of Ru catalysts supported on 600C12A7:OH-, 700C12A7:OH- and 800C12A7:OH- are omitted for readability. From Figure 2a one can see that Ru catalysts supported on C12A7:OHcalcined at different temperatures all have a distinctly higher CO2 methanation activity than Ru/-Al2O3, their catalytic activity (Table 2) varies inversely with the calcination temperature of 16

the supports except for the 400 oC-calcined sample, suggesting the surface area of the support is one factor affecting the catalytic activity. Figure S10 compares the catalytic performance of Ru supported on 500C12A7:OH-, -Al2O3 and a physical mixture of 12CaO and 7Al2O3. These results exhibit that the crystalline structure of the mayenite plays an important role in the catalysis of CO2 methanation over Ru supported on C12A7:OH-, and that the superior catalytic

oo f

activity of the Ru/C12A7:OH- catalysts is not ascribed to the contribution from CaO phase,

which is an alkali adsorbent for CO2 contained in the supports (Figure S2). We will discuss that

pr

CO2 is mainly adsorbed and activated on the support in mechanism studies in section 3.4. The

e-

OH- encaged in mayenite support is proposed to promote the conversion of CO2 into carbonates and bicarbonates on the support, which should contribute to the higher activity of

Pr

Ru/C12A7:OH- catalyst as compared to that of Ru/12CaO:7Al2O3.

Since Yoshitake Toda et al. [36] have reported that the surface of C12A7:e- can activate and

na l

split CO2 into CO(ads) at room temperature due to a high concentration of encaged electrons in the near-surface region; However the desorption of CO(ads) can only occur at temperature above

ur

800 oC. Therefore we expected to support Ru nanoparticles on the surface of C12A7:e- to

Jo

achieve a highly active catalyst for CO2 methanation. When the supports are reduced into electrides, the catalytic activity of Ru/C12A7:e- unexpectedly drops to a level even worse than Ru/-Al2O3 as shown in Figure 2b. Comparing the intrinsic activity (TOF at 225 oC in Table 2 calculated based on the activities corresponding to a CO2 conversion of lower than 10 %) with the electron density (Figure 1b) of the C12A7:e- supported catalysts, we can see that the TOF 17

increases with the increasing of electron density of the electride supports. It suggests that the encaged electrons in mayenite electride supports contribute significantly to the catalysis of CO2 methanation over Ru particles. The unexpected low TOF of Ru supported on 800C12A7:e- and 900C12A7:e- might be caused by the O2- ions in the framework of the supports (Figure 1b), however, there is no experimental evidence supporting this viewpoint at present.

oo f

CO2 methanation on Ru catalysts is a well-known structure-sensitive reaction [41-44]. To isolate the contribution of the support to the TOF from that of Ru particle size, we measured the

pr

intrinsic activity of Ru catalysts having metal loading of 0.93, 2.85 and 3.54 wt% (Table 2)

e-

supported on 500C12A7:OH-. Figure 3 demonstrates the dependence of TOF on FE measured by CO-titration method. Fitting the data of TOF versus FE, we obtained an equation as following.

Pr

𝑇𝑂𝐹 𝐹𝐸 𝑙𝑜𝑔10 = −1.404𝑙𝑜𝑔10 + 0.693

(4)

We calculated the TOF of Ru/500C12A7:OH- should be around 0.080 s-1 at FE = 18.8, which is

na l

more than 3 times higher than that of the 1.97%Ru/1000C12A7:e- with the same FE. Although the nature of the support might influence the structure and size of Ru on it [41], we propose the

ur

assumption that the dependence of TOF on the size of Ru particles supported on different

Jo

supports follows the same trend (i.e. the lines in Figure 3 have the same slope) as shown in Figure 3. Based on this assumption, we can obtain similar equation (5) and (6) for 1000C12A7:eand -Al2O3 supported catalysts, respectively. 𝑇𝑂𝐹 𝐹𝐸 𝑙𝑜𝑔10 = −1.404𝑙𝑜𝑔10 + 0.169

(5)

𝑇𝑂𝐹 𝐹𝐸 𝑙𝑜𝑔10 = −1.404𝑙𝑜𝑔10 + 0.0509

(6) 18

Based on equation (6), one can calculate that the TOF of Ru/-Al2O3 at FE = 18.8 is 0.018 s-1, which is slightly lower than that of Ru/1000C12A7:e-. Therefore, the intercept of the above equations should reflect the contribution of the support to the TOF of Ru. Based on the above calculations, the TOF on Ru follows the sequence of Ru/500C12A7:OH- (0.080 s-1) > Ru/1000C12A7:e- (0.024 s-1) >> Ru/-Al2O3 (0.0023 s-1) at the

oo f

same FE of 18.8%. M. Kitano et al. [35] have proposed that electrons encaged in C12A7:e- are transferred to Ru metal, which causes a substantial lowering of the work function of Ru by

pr

raising Fermi level. Therefore, the activation of N≡N can be accelerated by enhancing the back-

e-

donation from Ru to N2 π*-antibonding orbital. We have expected that electrons encaged in C12A7:e- should accelerate CO2 methanation reaction on Ru in a similar way observed in the

Pr

activation of N2 and have a higher catalytic activity than Ru/C12A7:OH-.

na l

3.4 Reaction mechanism

To find out the reason responsible for the unexpected catalytic behavior of the Ru catalysts

ur

supported on C12A7:OH- and C12A7:e-, we selected Ru/500C12A7:OH- and Ru/1000C12A7:e-

Jo

to conduct temperature-programmed surface reaction (TPSR), CO2-TPD and operando FTIR investigations because: (a) the two catalysts are the most active catalyst in their respective categories, and (b) the encaged anions in 1000C12A7:e- support are mainly electrons (Figure 1b).

19

3.4.1 Temperature-Programmed Surface Reaction (TPSR) Study Figure 4a shows that the onset temperature for CO2 methanation on Ru/500C12A7:OH- is ~ 250 ℃. The production of CH4 from CO2 methanation over Ru/500C12A7:OH- increases with increasing temperature and reaches a maximum (about 0.77) at about 360 ℃. CO is not identified in the product up to 450 ℃. In the temperature range of 450 – 620 ℃, CO yield increases with

oo f

decreasing in CH4 yield and increasing in CO2 content, indicating the simultaneously occurring of CH4 reforming and RWGS. Above 620 ℃, the residual CO2 drops again due to the enhanced

pr

RWGS. In order to understand further the catalysis of CO2 methanation, we conducted TPSR of

e-

CO hydrogenation over Ru/500C12A7:OH- (Figure 4c). Methanation of CO occurs also on Ru/500C12A7:OH-, but with a rising onset temperature to 325 oC, 75 oC higher than CO2

Pr

methanation. The maximum yield of CH4 is identified at 380 oC, 20 oC higher than CO2 methanation. The different onset and peak temperatures observed in the TPSR profiles for CO

na l

and CO2 methanation over Ru/500C12A7:OH- suggests that raising CO partial pressure inhibits the activity of methane formation over Ru/500C12A7:OH-. Owing to the molar ratio of H2 and

ur

CO is 4:1, the CH4 yield rises to about 98 % at ~ 380 ℃. CO2 is observed above ~ 420 ℃, and its

Jo

content gradually increases with increasing of temperature up to ~ 560 ℃ and then drops slowly. Because CO content keeps constant but CH4 yield declines in the temperature range from 420 to 500 ℃, we deduce that CO2 comes from water gas shift (WGS) reaction. As the encaged OH- groups are replaced by electrons in the framework of mayenites, the Ru/1000C12A7:e- exhibits distinguishable TPSR profiles of methanation of CO2 (Figure 4b) or 20

CO (Figure 4d). The onset temperature for CO2 methanation on Ru/1000C12A7:e- (Figure 4b) is ~ 300 ℃, 50 oC higher than Ru/500C12A7:OH-. This result is consistent with the activity measurements (Figure 2). The peak yield of CH4 is observed at 390 oC, 30 oC higher than that of Ru/500C12A7:OH-. However, Ru/1000C12A7:e- and Ru/500C12A7:OH- give a totally same CO production profile in TPSR of CO2 methanation, including the onset temperature and the yields

oo f

at corresponding temperatures (Figures 4a and b), which tells that the supports influence tinily the catalysis of Ru for CH4 reforming and RWGS. The most significant difference is found in the

pr

TPSR profiles of CO hydrogenation over the two catalysts. A very little amount of CH4 and CO2

e-

is detected in the TPSR profiles of CO hydrogenation over Ru/1000C12A7:e- above 600 ℃. It shows that Ru/1000C12A7:e- dose not catalyze CO methanation reaction under investigated

Pr

conditions. This result suggests that the inhibition effect of CO partial pressure on the activity of

na l

Ru/1000C12A7:e- is more serious than that of Ru/500C12A7:OH-.

3.4.2 Temperature-Programmed Desorption of CO2 Study

ur

To investigate the role of Ru and supports in the activation of CO2, we compared the CO2-

Jo

TPD profiles for Ru/500C12A7:OH-, 500C12A7:OH-, Ru/1000C12A7:e-, and 1000C12A7:e(Figure 5). Data shown in Figures 5b and d reveal that both the supports (500C12A7:OH- and 1000C12A7:e-) can active and split CO2(ads) into CO(ads) and O(ads). The intensity of desorbed CO from 1000C12A7:e- is drastically higher than that from 500C12A7:OH-, and the desorption of CO(ads) from 1000C12A:e- starts at around 580 oC, ~100 oC lower than that from 500C12A7:OH-. 21

In addition, the desorption of O(ads) from 1000C12A7:e- start at 400 oC, ~260oC lower than that from 500C12A7:OH-. These different desorption behaviors for CO(ads) and O(ads) observed for the two supports demonstrate that electrons in the near-surface region of mayenite electride accelerate the conversion of CO2 to CO and ease the desorption of CO(ads) and O(ads). Since the temperature range and shape of desorption profile of CO(ads) and O(ads) from 1000C12A7:e- are

oo f

significantly different as shown in Figure 5d, we deduce that the splitting of CO2(ads) into CO(ads) and O(ads) occurs at temperatures lower than the observed desorption temperature of CO.

pr

We integrate the TPD profiles of CO2 and CO in Figure 5 and compile the results in Table S1.

e-

The amount of CO and CO2 adsorbed on unit surface area of 1000C12A7:e- is 28 times of that of 500C12A7:OH-. Additionally, the percentage of CO2 decomposed into CO on electride is 9 times

Pr

larger than that on 500C12A7:OH-. These results suggest that the basicity of 1000C12A7:e- is drastically stronger than that of 500C12A7:OH- and CO2 can be adsorbed and activated on both

na l

supports. The electron encaged in the mayenite framework have radically higher capability to enhance the adsorption and activation of CO2 on its surface as compared to OH- contained in

ur

C12A7:OH-.

Jo

Ru supported on 500C12A7:OH- lowers the onset desorption temperature of CO(ads) by 80 o

C (from 680 to 600 oC). Additionally, a weak and broad desorption peak of O(ads) in the

temperature range from 380 to 500 oC appears for Ru/500C12A7:OH- catalyst, suggesting O(ads) on the interface of Ru and support (It is reported that CO2 can not be dissociated on the surface of Ru particles without the help of H. [45-47]) desorbs drastically easier than that on 500C12A7:OH-. 22

A similar decrease in CO(ads) desorption temperature from 580 (Figure 5d) to 510 oC (Figure 5c) is observed for Ru/1000C12A7:e-. Supporting Ru increases the total adsorption amount of CO2 for both catalysts. Especially for the electride supported Ru catalyst, its adsorption capacity for CO2 is almost doubled by Ru supporting due to its enlarged surface area during Ru-loading. However, the decomposition percentage of adsorbed CO2 was slightly decreased by Ru supporting on

oo f

electride, which might be caused by the reduced electron density on the surface of the support because electrons are partially transferred to Ru particles. These results demonstrates that

pr

supporting Ru on 1000C12A7:e- accelerates the desorption of CO(ads) and lowers the CO2

e-

activation ability of electrides. The decomposition of CO2 into CO is not varied by Ru particles on

encaged in 500C12A7:OH-. 3.4.3 Operando DRIFTS Study

Pr

500C12A7:OH-, suggesting the Ru particles are not electronically influenced by the OH- anions

na l

For the sake of clarifying the aforementioned distinct catalytic behaviors of the Ru catalysts supported on mayenite supports having OH- or e- encaged, we carried out a series of operando

ur

DRIFTS measurements to identify the key surface intermediates relevant to CO2 and CO

Jo

methanation on the two types of catalysts. We first recorded the operando DRIFT spectra (Figure 6) at steady state reactions of CO2

(Figures 6a and b) or CO (Figures 6c and d) methanation over Ru/500C12A7:OH- (Figures 6a and c) or the 500C12A7:OH- support (Figures 6b and d) at 150 – 350 ℃ and 0.1MPa. In Figures 6a and c, an IR absorption feature located at 1800 – 2100 cm-1 is observed from 150 ℃. We can 23

not find the same feature on the 500C12A7:OH- support (Figures 6b and d). It implies that this absorption feature is caused by adsorbates on Ru. The main band at 2016 (Figure 6a) or 2026 (Figure 6c) cm-1 should be ascribed to linearly chemisorbed CO(ads) on Ru (Ru-CO) [48,49]. The intensity of the CO(ads) band decreases with a shift of its peak position from 2016 (2026) to 1998 (2007) cm -1 as reaction temperature rises from 150 to 350 ℃. It has been identified that the

oo f

C-O stretching band shifts to lower frequency as the surface coverage of CO(ads) decreases [50]. The aforementioned redshift of CO adsorption band implies that the coverage of CO(ads) on the

pr

surface of Ru particles decreases with increasing temperature. Based on literatures [48,49,51],

e-

the shoulder at 2042 and 2063 cm-1 observed respectively in CO2 (Figure 6a) and CO (Figure 6c) hydrogenation on Ru/500C12A7:OH- has two plausible assignments: one is ascribed to the

Pr

absorption of tricarbonyl bonded on partially oxidized Ru+ (Ru+-(CO)3) (another stretching frequency of Ru+-(CO)3 might be 2115 cm-1); the other is ascribed to the absorption of

na l

dicarbonyl on metallic Ru (Ru-(CO)2) (another stretching frequency of Ru-(CO)2 should be 1962 or 1965 cm-1). Considering that at our reaction conditions the Ru particles on 500C12A7:OH- are

ur

reduced in metallic state, which is proved by XPS measurements will be discussed later, and that

Jo

the 2115 cm-1 band is absent in CO2 hydrogenation stream (Figure 6a), thus we suppose that the two shoulders at 2042 (2063) and 1962 (1965) cm-1 are related to Ru-(CO)2 and that the bands at 2180 and 2115 cm-1 observed in CO hydrogenation stream on Ru/500C12A7:OH- (Figure 6c) are absorptions of gas phase CO [52]. The shoulder at 1895 or 1903 cm-1 is assigned to bridged CO on Ru (Ru2CO) [49,53,54]. The gas-phase CO bands at around 2100 cm-1 do not appear in Figure 24

6a, in agreement with the result that CO was not detected under 450 ℃ in the corresponding TPSR experiment (Figure 4a). It reveals that the CO(ads) molecules on Ru particles do not desorb under 450 ℃. The peaks at 1304 and 3016 cm-1 associated to rotational fine structures of gas-phase CH4 appear above 200 ℃ [55,56]. The intensities of these two bands increase quickly, which is accompanied by the simultaneous reduction in CO(ads) coverage on Ru, as raising

oo f

reaction temperatures. It illustrates that CO(ads) on Ru particles should be an important reaction intermediate for CO2 methanation on Ru/C12A7:OH-.

pr

The absorption bands at frequencies lower than 1800 cm-1 are mainly assignable to

e-

carbonates, bicarbonates and formates adsorbed on the support [46,49]. Three groups of bands at (~1608, ~1487, 1226 cm-1), (1556, 1338, 1052 cm-1), and (~1780, 1153 cm-1) are respectively

Pr

assigned to surface bicarbonate, monocarbonate, and bridging carbonate [57,58]. These IR absorption features are identified under all the studied reaction conditions in Figure 6. We

na l

subtracted the IR spectrum of the 500C12A7:OH- support from that of the corresponding Ru/500C12A7:OH- at 150, 250, and 350 oC (Figure S11). The fact that the bands of carbonate

ur

and bicarbonate are not identified in the subtracted spectra at 150 and 250 oC suggests that these

Jo

adsorbates are mainly on the surface of the 500C12A7:OH- support. However, a few broad bands in the region of 1000 to 1750 cm-1 can be observed in the subtracted spectrum at 350 oC, which might be caused by carbons deposited on the catalyst [59]. The IR absorptions at frequencies in the range of 2450 – 2700 cm-1 and of 2800 – 2970 cm-1 in Figures 6b and d are assignable to adsorbed formates on the support [60,61]. These formate bands are not observed in Figures 6a 25

and b. It means that supporting Ru accelerates the conversion of these formates produced on the support, suggesting that formates formed on the mayenite are one of the intermediate species for CO2 methanation over Ru/500C12A7:OH-. In Figure 6b, the peaks at 2180 and 2115 cm-1 related to gas-phase CO [52,61] appear at 300 ℃ and their intensities increase with increasing temperature. The gas-phase CO should be formed via RWGS reaction catalyzed by the

oo f

500C12A7:OH- support, which is accompanied by the concurrent appearance of a tiny CH4 peak at 3016 cm-1.

pr

From Figures 6a and b, we suppose Ru-CO(ads), which is supposed being the product of

e-

formate reduction [9, 62], is one of the key intermediate for CO2 methanation over Ru/500C12A7:OH-; thus, we expected that this catalyst should have an improved performance

Pr

for CO methanation. However, the appearance of methane band is surprisingly delayed to 300 ℃ (Figure 6c), 100 oC higher than CO2 methanation. As a result of higher CO coverage on Ru

na l

particles, the intensity of the linearly adsorbed CO(ads) is obviously stronger than that observed in CO2 methanation (Figure 6a), and its peak shifted from 2026 to 2009 cm-1 as reaction

ur

temperature rises from 150 to 350 oC. This observed redshift suggests that the CO surface

Jo

coverage decreases with increasing of CO methanation rate. The increased CO(ads) coverage on Ru surface should decrease the available Ru sites for H(ads), which reacts with CO(ads) to form methane; consequently, a lower catalytic activity for CO methanation on Ru/500C12A7:OH- is observed. These results also demonstrates that the conversion of Ru-CO(ads) to CH4 is not the rate determining steps for CO2 methanation on Ru/C12A7:OH- under our reaction conditions. 26

Figure 7 shows the DRIFT spectrum of CO2 (a, b) or CO (c, d) methanation over Ru/1000C12A7:e- catalyst (a, c) or 1000C12A7:e- support (b, d). Comparing Figures 6a and 7a, the peak at 2016 cm-1 related to linearly adsorbed CO(ads) on Ru/1000C12A7:e- is significantly weaker than that on Ru/500C12A7. We suppose that the reason is the CO(ads) dissociates further into C(ads) and O(ads), which do not absorb IR, on Ru particles, which accept electrons from the

oo f

1000C12A7:e- substrate. The proofs for this point will be discussed in detail later. The peaks due to carbonates and bicarbonate below 1800 cm-1 appears on Ru/1000C12A7:e- (Figure 7a) and

pr

1000C12A7:e- (Figure 7b), but almost cannot be identified in the subtracted IR spectra (Figure

e-

S12). However, the CO methanation does not occur on Ru/1000C12A7:e- at temperatures below 500 ℃ (Figure 7c), which agrees with the TPSR results (Figure 4d). Comparing Figures 7b and d,

Pr

we find no significant differences other than the peak related to gas-phase CO and CO2. It suggests that CO2 is readily converted into CO on the electride surface, which is supported by

na l

the high conversion of CO2 into CO (86.8%) (Table S1) calculated from CO2-TPD profiles (Figure 5).

ur

To verify the assumption that CO(ads) dissociates into C(ads) on Ru/1000C12A7:e-, we varied the CO partial pressure via increasing its flow rate from 1 to 10 mL min-1 while keeping

Jo

that of H2 and Ar constant and collected the DRIFT spectrum over Ru/1000C12A7:e- at 350 ℃ and 0.1 MPa (Figure 8). At low CO partial pressure, i.e. introducing 1 mL min-1 of CO for 40 min (spectra in black in Figure 8), the bands ascribed to adsorbed CO(ads) and to gas-phase methane are identified and their intensities increase in the first 10 min and then decrease in the 27

last 10 min. The reduction in gas-phase methane and enrichment in gas-phase CO in the last 10 min suggest that the catalytic activity decreases gradually with the proceeding of CO methanation. As we raise the CO flow rate to 2.5 (spectra in red in Figure 8) and 5 (spectra in purple in Figure 8) mL min-1, the intensity of the bands ascribable to adsorbed CO(ads) and to gas-phase CH4 simultaneously decreases in a similar trend, suggesting the deactivation of

oo f

Ru/1000C12A7:e- becomes more serious. The peaks related to CH4 and adsorbed CO(ads)

disappear when the flow rate of CO rises to 10 mL min-1 (spectra in blue in Figure 8). This

pr

experiment confirms that the CO(ads) works as vital intermediate in the methanation of CO and

e-

CO2 on the surface of Ru/1000C12A7:e- and high CO partial pressure accelerates the deactivation of the catalyst. Subsequent to the confirmation of the effect of CO partial pressure

Pr

(Figure 8), we tried to treat the catalyst with 30 mL min-1 of 5 % O2 in Ar at 350 ℃ for an hour and to monitor the effluent with a mass spectrometer (Figure 9a). The formation of CO2 from

na l

oxidation of deposited carbons is confirmed (Figure 9a). Then, we reduced the oxidized Ru particles in 50 mL min-1 of 4H2 +1Ar at 350 oC for an hour, and subsequently verified the activity

ur

of CO methanation over the reactivated Ru/1000C12A7:e- catalyst by introducing 1 mL min-1 of

Jo

CO into the system (Figure 9b). It can be seen from Figure 9b that the catalyst exhibits similar catalytic performance for CO methanation as fresh one (Figure 8). Based on these results, we conclude that carbon deposited on Ru/1000C12A7:e- from CO(ads) at the reaction conditions of CO2 or CO methanation plays a vital role for their unexpected catalytic performances observed. In order to investigate the effect of the mayenite supports encaged with OH- or e- on the 28

catalytic behavior of CO2 or CO methanation on Ru particles, we characterized the valence state of Ru on the two kinds of mayenite supports with XPS (Figure 10). The samples were reduced in H2 at 300 oC for 3 h, pressed into disks and sealed in sample vials in Ar atmosphere in a glove box. The disks were transferred into the XPS analytic chamber as quick as possible. The Ru 3d5/2 peak of the original surface of the Ru/500C12A7:OH- is located at 280.8 eV, which suggests that

oo f

the Ru particles on the surface of the disk were oxidized during the transferring of the sample into the chamber. However, the corresponding Ru 3d5/2 peak of the original surface of

pr

Ru/1000C12A7:e- is observed at 279.5 eV, 0.8 eV lower than the binding energy of metallic Ru

e-

(280.2 eV). This result reveals that Ru/1000C12A7:e- has higher resistance to oxidation as compared to Ru/500C12A7:OH-. After the samples were sputtered with Ar ions for 20 and 50 s,

Pr

the XPS spectra of the two catalysts were measured again (Figures 10b and d, the spectra recorded at 50 s are the same with those recorded at 20 s). The 3d5/2 peak of Ru on

na l

500C12A7:OH- shifts to 280.2 eV, indicating the Ru particles on mayenite encaged with OH- are at metallic state. On the other hand, the 3d5/2 peak of Ru on 1000C12A7:e- is located at 279.0 eV,

ur

suggesting the electrons encaged in C12A7:e- are transferred to the Ru particles and the Ru

Jo

particles are negatively charged. The transferring of electrons encaged in C12A7:e- to the Ru particles causes a significantly lowering of the work function of Ru by raising the Fermi level. As a result, back donation from Ru with an excess electron density to the CO(ads) antibonding orbital is enhanced, which leads to a weakening of the C=O bond of adsorbed CO(ads) and to an accelerating of dissociation of CO(ads) into C(ads) and O(ads). The C(ads) 29

deposits and blocks the subsequent methanation reactions. In summary, we propose the reaction mechanism of CO2 methanation over Ru supported on mayenites as Scheme 1. Since the binding of CO2 at the surface of Ru (Eads (Ru-CO2) = -0.52 eV) is drastically weaker than that of CO (Eads (Ru-CO) = -2.30 eV) [9], thus the activation of CO2 directly on Ru surface for supported catalysts is not favorable. For hydrogenation of CO2 on

oo f

supported Ru, Pd, Ni catalyst at atmospheric pressure, the CO2 is generally activated on the

supports [9,63,64]. Here we suppose that the gas-phase CO2(g) molecules are first adsorbed on

pr

the mayenite supports (R1) with the help of encaged OH- and e- to produce the surface carbonates

e-

and bicarbonates, which are hydrogenated into formates on the interface between the support and Ru (R2). Then the formates dissociate into *CO and *OH with the help of *H on Ru (R3), the

Pr

adsorbed *CO can be converted by *H atoms on the catalyst into CH4 (R4) or dissociated into *C (R5). The relative reaction rate of R4 and R5 on Ru particles at different charge states plays a key

na l

role in the different catalytic behavior of Ru supported on mayenites encaged with different anions (OH- or e-). The Ru particles accepting electrons from the electride substrate drastically

ur

accelerate R5. As a consequence, *C forms quickly and converts into deposited carbons (*Cn),

Jo

which can hinder the methanation process due to the relatively slower conversion rates (R6 and R8) as compared to R4.

3.5 Catalytic stability of Ru/C12A7:OH- and Ru/C12A7:eThe catalytic stability of the catalysts were evaluated by confirming the change in their 30

intrinsic activity (at around 20 % conversion of CO2) after 72 h on stream tests at their corresponding maximum reaction rates. Figure 11a shows that the maximum conversion of CO2 is almost constant after a slight initial reduction and the intrinsic conversion of CO2 over Ru/500C12A7:OH- at 275 ℃ decreases from 26 % to 22 %. On the other hand, the CO2 conversion decreases gradually from 74 % to 40 % over Ru/1000C12A7:e- with its intrinsic

oo f

activity dropping from 20 % to 5 %. Besides the CaO phase is converted into CaCO3 phase

during the on stream tests, the XRD (Figure S13) shows no obvious changes. A slight growth of

pr

Ru particles on both substrates is observed as shown in Figure S14, which should be responsible

e-

for the deactivation identified for Ru/500C12A7:OH-. The Ru particles on 1000C12A7:e- being surrounded by a layer of carbons are observable in Figure S14d.

Pr

Figure 12 compares the UV-Visible absorption spectra of Ru/500C12A7:OH- and Ru/1000 C12A7:e−. The disappearance of the absorption band at around 2.8 eV and the optical absorption

na l

edge appearance at around 5.3 eV after on-stream test of Ru/1000 C12A7:e− reveal that the

ur

encaged electrons are replaced by OH- ions.

Jo

4. Conclusions

Mayenites encaged with different anions (OH- or e-) have been synthesized and applied as

supports for Ru catalysts in CO2 methanation for the first time. The Ru/C12A7:OH- catalyst exhibit a significantly better activity (TOF at 225 oC = 6.04  10-2 s-1, Ea = 70.9 kJ mol-1) than that (TOF at 225 oC = 0.23  10-2 s-1 , Ea = 105.8 kJ mol-1) of Ru/-Al2O3. The catalytic 31

performance of Ru/C12A7:e- is unexpectedly inferior to that of Ru/C12A7:OH-. Based on operando DRIFTS investigations, we propose a reaction mechanism for CO2 methanation on Ru supported on mayenites. It has been found that the OH- or e- encaged in mayenite promotes the formation of carbonates and bicarbonates from CO2 on the supports. It is supposed that at the interface between Ru and mayenite support the surface carbonates and bicarbonates are

oo f

hydrogenated into formates, which are further dissociated into *CO adsorbed on Ru. The formed *CO is a key intermediate for CO2 methanation and carbon deposition. The different anions (OH-

pr

or e-) contained in the cages of mayenite supports influences drastically the ionization potential

e-

of Ru. Electrons encaged in C12A7:e- are transferred to Ru metal, thereby enhancing the back donation from Ru to the antibonding orbitals of *CO. As a result, the dissociation of *CO

Pr

intermediate into *C and subsequent into deposition of carbons, which is responsible for the inferior catalytic performance, is accelerated significantly on Ru/C12A7:e- catalysts. The

na l

experimental results suggest that changing encaged anion species in mayenite substrates is an

ur

effective technique for tailoring the catalytic activity of metallic sites supported on them.

Jo

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China

(21573163) and the Natural Science Foundation of Hubei Province (2015CFA017).

32

References [1] Earth's CO2 Home Page. https://www.co2.earth/co2-ice-core-data (accessed Jan 10, 2019). [2] J.P. Smol, Climate Change: A planet in flux, Nature 483 (2012) S12-S15.

oo f

[3] M. Mikkelsen, M. Jørgensen, F.C. Krebs, The teraton challenge. A review of fixation and transformation of carbon dioxide, Energy Environ. Sci. 3 (2010) 43-81.

pr

[4] H.S. de Boer, L. Grond, H. Moll, R. Benders, The application of power-to-gas, pumped hydro

penetration levels, Energy 72 (2014) 360-370.

e-

storage and compressed air energy storage in an electricity system at different wind power

Pr

[5] M. Götz, J. Lefebvre, F. Mörs, A. McDaniel Koch, F. Graf, S. Bajohr, R. Reimert, T. Kolb, Renewable Power-to-Gas: A technological and economic review, Renew. Energ. 85 (2016)

na l

1371-1390.

[6] K. Li, A. Cousins, H. Yu, P. Feron, M. Tade, W. Luo, J. Chen, Systematic study of aqueous

ur

monoethanolamine-based CO2 capture process: model development and process improvement, Energ. Sci. Eng. 4 (2016) 23-39.

Jo

[7] P. M. Bravo, D. P. Debecker, Combining CO2 capture and catalytic conversion of methane, Waste Dispo. Sust. Energ. 1 (2019) 53-65. [8] W. Wang, S. Wang, X. Ma, J. Gong, Recent advances in CO2 capture and utilization, Chem. Sus. Chem. 40 (2011) 3703-3727. [9] X. Wang, H. Shi, J.H. Kwak, J. Szanyi, Mechanism of CO2 Hydrogenation on Pd/Al2O3 33

Catalysts: Kinetics and Transient DRIFTS-MS Studies, ACS Catal. 5 (2015) 6337-6349. [10] A. Quindimil, U. De-La-Torre, B. Pereda-Ayo, J.A. González-Marcos, J.R. GonzálezVelasco, Ni catalysts with La as promoter supported over Y- and BETA- zeolites for CO2 methanation, Appl. Catal. B: Environ. 238 (2018) 393-403. [11] J. Lin, C. Ma, Q. Wang, Y. Xu, G. Ma, J. Wang, H. Wang, C. Dong, C. Zhang, M. Ding,

oo f

Enhanced low-temperature performance of CO2 methanation over mesoporous Ni/Al2O3-ZrO2 catalysts, Appl. Catal. B: Environ. 243 (2019) 262-272.

pr

[12] J.A.H. Dreyer, P. Li, L. Zhang, G.K. Beh, R. Zhang, P.H.L. Sit, W.Y. Teoh, Influence of the

e-

oxide support reducibility on the CO2 methanation over Ru-based catalysts, Appl. Catal. B: Environ. 219 (2017) 715-726.

Pr

[13] M.A.A. Aziz, A.A. Jalil, S. Triwahyono, R.R. Mukti, Y.H. Taufiq-Yap, M.R. Sazegar, Highly active Ni-promoted mesostructured silica nanoparticles for CO2 methanation, Appl.

na l

Catal. B: Environ. 147 (2014) 359-368.

[14] D. Beierlein, D. Häussermann, M. Pfeifer, T. Schwarz, K. Stöwe, Y. Traa, E. Klemm, Is the

ur

CO2 methanation on highly loaded Ni-Al2O3 catalysts really structure-sensitive, Appl. Catal. B:

Jo

Environ. 247 (2019) 200-219.

[15] A. Kim, D. P. Debecker, F. Devred, V. Dubois, C. Sanchez, C. Sassoye, CO2 methanation on Ru/TiO2 catalysts: On the effect of mixing anatase and rutile TiO2 Supports, Appl. Catal. B: Environ. 220 (2018) 615-625. [16] S. Sharma, Z. Hu, P. Zhang, E.W. McFarland, H. Metiu, CO2 methanation on Ru-doped 34

ceria, J. Catal. 278 (2011) 297-309. [17] A. Beuls, C. Swalus, M. Jacquemin, G. Heyen, A. Karelovic, P. Ruiz, Methanation of CO2: Further insight into the mechanism over Rh/γ-Al2O3 catalyst. Appl. Catal. B: Environ. 113-114 (2012) 2-10. [18] C. Li, S. Zhang, B. Zhang, D. Su, S. He, Y. Zhao, J. Liu, F. Wang, M. Wei, D.G. Evans, X.

oo f

Duan, Photohole-oxidation-assisted anchoring of ultra-small Ru clusters onto TiO2 with excellent catalytic activity and stability, J. Mater. Chem. A 1 (2013) 2461-2467.

pr

[19] G. Garbarino, P. Riani, L. Magistri, G. Busca, A study of the methanation of carbon dioxide

e-

on Ni/Al2O3 catalysts at atmospheric pressure, Int. J. Hydrogen Energy 39 (2014) 11557-11565. [20] P. Panagiotopoulou, Methanation of CO2 over alkali-promoted Ru/TiO2 catalysts: II. Effect

Pr

of alkali additives on the reaction pathway, Appl. Catal. B: Environ. 236 (2018) 162-170. [21] J.-N. Park, E.W. McFarland, A highly dispersed Pd–Mg/SiO2 catalyst active for

na l

methanation of CO2, J. Catal. 266 (2009) 92-97. [22] T. Riedel, M. Claeys, H. Schulz, G. Schaub, S.-S. Nam, K.-W. Jun, M.-J. Choi, G. Kishan,

ur

K.-W. Lee, Comparative study of Fischer–Tropsch synthesis with H2/CO and H2/CO2 syngas

Jo

using Fe- and Co-based catalysts, Appl. Catal. A: Gen. 186 (1999) 201-213. [23] J. Kirchner, J.K. Anolleck, H. Lösch, S. Kureti, Methanation of CO2 on iron based catalysts, Appl. Catal. B: Environ. 223 (2018) 47-59. [24] S. Scirè, C. Crisafulli, R. Maggiore, S. Minicò, S. Galvagno, Influence of the support on CO2 methanation over Ru catalysts: an FT-IR study, Catal. Lett. 51 (1998) 41-45. 35

[25] J. Dou, Y. Sheng, C. Choong, L. Chen, H.C. Zeng, Silica nanowires encapsulated Ru nanoparticles as stable nanocatalysts for selective hydrogenation of CO2 to CO, Appl. Catal. B: Environ. 219 (2017) 580-591. [26] F. Wang, S. He, H. Chen, B. Wang, L. Zheng, M. Wei, D.G. Evans, X. Duan, Active Site

Soc. 138 (2016) 6298-6305.

oo f

Dependent Reaction Mechanism over Ru/CeO2 Catalyst toward CO2 Methanation, J. Am. Chem.

[27] H. Song, J. Yang, J. Zhao, L. Chou, Methanation of Carbon Dioxide over a Highly

pr

Dispersed Ni/La2O3 Catalyst, Chin. J. Catal. 31 (2010) 21-23.

e-

[28] S. Mori, W.C. Xu, T. Ishidzuki, N. Ogasawara, J. Imai, K. Kobayashi, Mechanochemical activation of catalysts for CO2 methanation, Appl. Catal. A: Gen. 137 (1996) 255-268.

Pr

[29] S. Matsuishi, Y. Toda, M. Miyakawa, K. Hayashi, T. Kamiya, M. Hirano, I. Tanaka, H. Hosono, High-density electron anions in a nanoporous single crystal: [Ca24Al28O64]4+(4e-),

na l

Science 301 (2003) 626-629.

[30] S.W. Kim, S. Matsuishi, T. Nomura, Y. Kubota, M. Takata, K. Hayashi, T. Kamiya, M.

ur

Hirano, H. Hosono, Metallic State in a Lime−Alumina Compound with Nanoporous Structure,

Jo

Nano Lett. 7 (2007) 1138-1143. [31] J.A. Imlach, L.S.D. Glasser, F.P. Glasser, Excess oxygen and the stability of “12CaO.7A12O3”, Cem. Concr. Res. 1 (1971) 57-61. [32] K. Hayashi, M. Hirano, H. Hosono, Thermodynamics and kinetics of hydroxide ion formation in 12CaO · 7Al2O3, J. Phys. Chem. 109 (2005) 11900-11906. 36

[33] S. Matsuishi, K. Hayashi, M. Hirano, H. Hosono, Hydride ion as photoelectron donor in microporous crystal, J. Am. Chem. Soc. 127 (2005) 12454-12455. [34] K. Hayashi, P.V. Sushko, D.M. Ramo, A.L. Shluger, S. Watauchi, I. Tanaka, S. Matsuishi, M. Hirano, H. Hosono, Nanoporous crystal 12CaO.7Al2O3: a playground for studies of ultraviolet optical absorption of negative ions, J. Phys. Chem. B 111 (2007) 1946-1956.

oo f

[35] M. Kitano, Y. Inoue, Y. Yamazaki, F. Hayashi, S. Kanbara, S. Matsuishi, T. Yokoyama, S.W.

reversible hydrogen store, Nat. Chem. 4 (2012) 934-940.

pr

Kim, M. Hara, H. Hosono, Ammonia synthesis using a stable electride as an electron donor and

e-

[36] Y. Toda, H. Hirayama, N. Kuganathan, A. Torrisi, P.V. Sushko, H. Hosono, Activation and splitting of carbon dioxide on the surface of an inorganic electride material, Nat. Commun. 4

Pr

(2013) 2378.

[37] C. Swalus, M. Jacquemin, C. Poleunis, P. Bertrand, P. Ruiz, CO2 methanation on Rh/γ-

na l

Al2O3 catalyst at low temperature: “In situ” supply of hydrogen by Ni/activated carbon catalyst, Appl. Catal. B: Environ. 125 (2012) 41-50.

ur

[38] C. Li, D. Hirabayashi, K. Suzuki, Synthesis of higher surface area mayenite by

Jo

hydrothermal method, Mater. Res. Bull. 46 (2011) 1307-1310. [39] S. Matsuishi, T. Nomura, M. Hirano, K. Kodama, S.-i. Shamoto, H. Hosono, Direct Synthesis of Powdery Inorganic Electride [Ca24Al28O64]4+(e–)4 and Determination of Oxygen Stoichiometry, Chem. Mater. 21 (2009) 2589-2591. [40] A.B. And, M. Bonarowska, Relation between Crystallite Size and Dispersion on Supported 37

Metal Catalysts, Langmuir 13 (1997) 5613-5620. [41] M. Che, C.O. Bennett, The influence of particle size on the catalytic properties of supported metals, Adv. Catal. 36 (1989) 55-172. [42] D.L. King, A Fischer-Tropsch study of supported ruthenium catalysts, J. Catal. 51 (1978) 386-397.

oo f

[43] C.S. Kellner, A.T. Bell, Effects of dispersion on the activity and selectivity of aluminasupported ruthenium catalysts for carbon monoxide, J. Catal. 75 (1982) 251-261.

pr

[44] P. Panagiotopoulou, D.I. Kondarides, X.E. Verykios, Selective methanantion of CO over

e-

supported Ru catalysts, Appl. Catal. B: Environ. 88 (2009) 470-478.

Catal. A: Gen. 542 (2017) 63-70.

Pr

[45] P. Panagiotopoulou, Hydrogenation of CO2 over supported noble metal catalysts, Appl.

[46] X. Wang, Y. Hong, H. Shi, J. Szanyi, Kinetic modeling and tansient DRIFT-MS studies of

na l

CO2 methanation over Ru/Al2O3 catalysts, J. Catal. 343 (2016) 185-195. [47] F. Solymosi, A. Erdöhelyi, T. Bánsági, Methanation of CO2 on supported rhodium catalyst,

ur

J. Catal. 68 (1981) 371-382.

Jo

[48] S.Y. Chin, C.T. Williams, M.D. Amiridis, FTIR studies of CO adsorption on Al2O3- and SiO2-supported Ru catalysts, J. Phys. Chem. B 110 (2006) 871-872. [49] S. Eckle, Y. Denkwitz, R.J. Behm, Activity, selectivity, and adsorbed reaction intermediates/reaction side products in the selective methanation of CO in reformate gases on supported Ru catalysts, J. Catal. 269 (2010) 255-268. 38

[50] H. Pfnür, D. Menzel, F.M. Hoffman, A.M. Bradshaw, High resolution vibtational spectrocopy of CO on Ru(001) the importance of lateral interactions, Surf. Sci. 93 (1980) 431452. [51] J.L. Robbins, Chemistry of supported Ru:CO-induced oxidation of Ru at 310 K, J. Catal. 115 (1989) 120-131.

oo f

[52] J. Aβmann, E. Löffler, A. Birkner, M. Muhler, Ruthenium as oxidation catalyst: bridging the pressure and material gaps between ideal and real systems in heterogeneous catalysis by

pr

applying DRIFT spectroscopy and the ATP reactor, Catal. Today 85 (2003) 235-249.

e-

[53] H. Yamasaki, Y. Kobori, S. Naito, T. Onishi, K. Tamaru, Infrared study of the reaction of H2 + CO on Ru/SiO2 catalyst, J. Chem. Soc. Faraday Trans. 1, 77 (1981) 2913-2925.

Pr

[54] C. Elmasides, D. I. Kondarides, a. W. Grünert, X. E. Verykios, XPS and FTIR Study of Ru/Al2O3 and Ru/TiO2 Catalysts: Reduction Characteristics and Interaction with a

na l

Methane−Oxygen Mixture, J. Phys. Chem. B 103 (1999) 5227-5239. [55] S. Sharma, K.B. Sravan Kumar, Y.M. Chandnani, V.S. Phani Kumar, B.P. Gangwar, A.

ur

Singhal, P.A. Deshpande, Mechanistic Insights into CO2 Methanation over Ru-Substituted CeO2,

Jo

J. Phys. Chem. C 120 (2016) 14101-14112. [56] C. Schild, A. Wokaun, A. Baiker, On the mechanism of CO and CO2 hydrogenation reactions on zirconia-supported catalysts: a diffuse reflectance FTIR study: Part I. Identification of surface species and methanation reactions on palladium/zirconia catalysts, J. Mol. Catal. 63 (1990) 223-242. 39

[57] R. Philipp, K. Fujimoto, FTIR spectroscopic study of CO2 adsorption/desorption on MgO/CaO catalysts, J. Phys. Chem. 96 (1992) 9035-9038. [58] A.M. Turek, I.E. Wachs, E. Decanio, Acidic properties of alumina-supported metal oxide catalysts: an infrared spectroscopy study, J. Phys. Chem. 96 (1992) 5000-5007. [59] V. Pawar, D. Ray, C. Subrahmanyam, V.M. Janardhanan, Study of Short-Term Catalyst

oo f

Deactivation Due to Carbon Deposition during Biogas Dry Reforming on Supported Ni Catalyst, Energy Fuels 29 (2015) 8047-8052.

pr

[60] S. Kouva, K. Honkala, L. Lefferts, J. Kanervo, Review: monoclinic zirconia, its surface sits

e-

and their interaction with carbon monoxide, Catal. Sci. Technol. 5 (2015) 3473-3490. [61] T. Shido, Y. Iwasawa, Regulation of reaction intermediate by reactant in the water-gas shift

Pr

reaction on CeO2, in relation to reactant-promoted mechanism, J. Catal. 136 (1992) 493-503. [62] S.T. Zhang, H. Yan, M. Wei, D.G. Evans, X. Duan, Hydrogenation mechanism of carbon

na l

dioxide and carbon monoxide on Ru(0001) surface: a density functional theory study, RSC Adv. 4 (2014) 30241-30249.

ur

[63] J.H. Hwak, L. Kovarik, J. Szanyi, CO2 reduction on supported Ru/Al2O3 catalysts: cluster

Jo

size dependence of product selectivity, ACS Catal. 3 (2013) 2449-2455. [64] P.A. Ussa Aldana, F. Ocampo, K. Kobl, B. Louis, F. Thibault-Starzyk, M. Daturi, P. Bazin, S. Thomas, A.C. Roger, Catalytic CO2 valorization into CH4 on Ni-based ceria-zirconia. Reaction mechansim by operando IR spectroscopy, Catal. Today 215 (2013) 201-207.

40

oo f pr ePr

Figure 1. UV-Visible absorption spectra (obtained by Kubelka-Munk transformation of diffuse

na l

reflectance spectra) of (a) C12A7:OH- and (b) C12A7:e- samples. The calculated electron densities

Jo

ur

of C12A7:e- samples were also listed in (b).

41

42

na l

ur

Jo

oo f

pr

e-

Pr

Figure 2. Catalytic performance of Ru catalysts supported on C12A7:OH-, -Al2O3, or C12A7:efor CO2 methanation: (a and b) CO2 conversion and (c and d) CH4 yield. Reaction conditions: 50.0

Jo

ur

na l

Pr

e-

pr

oo f

mg catalyst diluted in 0.50 g quartz sand, 60 mL min-1 reactant gas (H2:CO2:Ar = 4:1:1), 0.1 MPa.

43

Figure 3. Turnover frequency (TOF) as a function of fraction exposed (FE) for Ru catalysts supported on 500C12A7:OH-, -Al2O3 and 1000C12A7:e-. Closed symbols are experimental data, open symbols are calculated TOF values by assuming that the dependence of TOF on Ru particle

oo f

size obeys equations with the same slope and the contribution of substrate is reflected by the

Jo

ur

na l

Pr

e-

pr

intercept of the equation.

44

oo f pr

e-

Figure 4. Temperature-programmed surface reaction (TPSR) between H2 and (a and b) CO2 or (c

Pr

and d) CO over (a and c) Ru/500C12A7:OH- or (b and d) Ru/1000C12A7:e-. Reaction conditions: 100.0 mg catalyst, 60 mL min-1 of reactant gas (H2 : CO2 or CO : Ar = 4 : 1 : 1). The partial pressure

na l

of gases contained in the product was normalized based on Ar. The catalysts were reduced in 60

Jo

ur

mL min-1 of 2H2 + 1Ar at 300 ℃for 5 h.

45

oo f pr ePr na l

Figure 5. Temperature-programmed desorption (TPD) profiles of CO2 over (a) Ru/500C12A7:OH-, (b)

ur

500C12A7:OH-, (c) Ru/1000C12A7:e-, and (d) 1000C12A7:e-. The catalysts (100.0 mg) were first reduced in

Jo

4H2 + 2Ar (60 mL min-1) at 300 oC for 4 h, and then purged with 20 mL min-1 Ar at 300 oC for 30 min. The adsorption of CO2 was carried out in 10 mL min-1 CO2 at 50 oC for 1 h. After purging with 10 mL min-1 Ar for 3 h, the catalysts were heated from 50 to 900 oC at 5 oC min-1 in a flow of 10 mL min-1 Ar.

46

oo f pr ePr na l ur Jo

Figure 6. In-situ DRIFT spectra collected during methanation reaction over (a and c) Ru/500C12A7:OH- catalyst and (b and d) 500C12A7:OH- support at 150-350 ℃in 60 mL min-1 of (a and b) 4H2 + 1CO2 +1Ar or (c and d) 4H2 + 1CO + 1Ar. The samples were pre-reduced in 50 mL min-1 of 4H2 + 1Ar at 300 ℃for 5 h. 47

48

na l

ur

Jo

oo f

pr

e-

Pr

Figure 7. In-situ DRIFT spectra collected during methanation reaction over (a and c) Ru/1000C12A7:e- catalyst and (b and d) 1000C12A7:e- support at different temperatures in 60 mL min-1 of (a and b) 4H2 + 1CO2 + 1Ar or (c and d) 4H2 + 1CO + 1Ar. The samples were pre-treated

Jo

ur

na l

Pr

e-

pr

oo f

in 50 mL min-1 of 4H2 + 1Ar at 300 ℃for 5 h.

49

oo f

pr

Figure 8. In-situ DRIFT spectra collected at 350 ℃ over Ru/1000C12A7:e- when raising the flow

e-

rate of CO from 1 to 2.5, 5 and 10 mL min-1, during which the flow rate of H2 and Ar was kept

Jo

ur

na l

+ 1Ar at 300 ℃for 5 h.

Pr

constant at 40 and 10 mL min-1, respectively. The catalyst was pre-reduced in 50 mL min-1 of 4H2

50

oo f pr e-

Figure 9. (a) MS signals of CO2 (m/z = 44) and O2 (m/z = 32) obtained during the reactivation of

Pr

the deactivated Ru/1000C12A7:e- in the transient CO methanation tests via feeding 30 mL min-1 of 5% O2 in Ar at 350 ℃. (b) In-situ DRIFT spectra collected after 1 mL min-1 CO + 40 mL min-1

Jo

ur

na l

H2 + 10 mL min-1 Ar were fed to the reactivated Ru/1000C12A7:e- catalyst at 350 oC.

51

oo f pr ePr na l

Figure 10. Ru 3d5/2 XPS spectra of (a, b) Ru/500C12A7:OH- and (c, d) Ru/1000C12A7:e-. The

ur

catalysts were pressed into disk after being reduced in H2 at 300 oC for 3 h and sealed in sample

Jo

vials in Ar atmosphere in a glove box, and the disks were transfer into XPS analytic chamber as quick as possible. (a, c) original surface of the reduced disks exposed air during transferring to analytic chamber, (b, d) fresh surface of the disks being etched for 20 s.

52

oo f pr ePr na l ur

Jo

Scheme 1. Proposed reaction mechanism for CO2 methanation over Ru/500C12A7:OH- or Ru/1000C12A7:e-.

53

54

na l

ur

Jo

oo f

pr

e-

Pr

Figure 11. CO2 conversion and CH4 yield over (a) Ru/500C12A7:OH- at 325 ℃(maximum CO2 conversion) and 275 oC (~20% CO2 conversion, detected before and after the time on-stream test

oo f

to confirm the change in intrinsic activity of the catalyst) and (b) Ru/1000C12A7:e− at 350 ℃ (maximum CO2 conversion) and 300 oC (~20% CO2 conversion, detected before and after the time

pr

on-stream test to confirm the change in intrinsic activity of the catalyst) vs time on stream.

Jo

ur

na l

Pr

e-

Reaction conditions: 100.0 mg catalyst, 60 mL min-1 of 4H2 + 1CO2 + 1Ar, 0.1 MPa.

55

oo f

Figure 12. UV-Vis absorption spectra (obtained by Kubelka-Munk transformation of diffuse reflectance spectra) of (a) 500C12A7:OH-, (b) fresh Ru/500C12A7:OH-, (c) reduced

pr

Ru/500C12A7:OH-, (d) aged Ru/500C12A7:OH-, (e) 1000C12A7:e−, (f) fresh Ru/1000C12A7:e−,

Jo

ur

na l

Pr

e-

(g) reduced Ru/1000C12A7:e-, and (h) aged Ru/1000C12A7:e−.

56

Table 1. Morphology of the synthesized C12A7:OH-1 and C12A7:e- supports BET surface area (m2 g-1)

400C12A7:OH-1

53

0.050

3.8

500C12A7:OH-1

42

0.066

6.3

600C12A7:OH-1

37

0.088

700C12A7:OH-1

32

0.091

800C12A7:OH-1

23

0.090

900C12A7:OH-1

17

0.084

19.5

-Al2O3

66

0.230 0.043

20.0

900C12A7:e-

11.5

pr

15.4

3

0.008

12.0

2

0.006

11.0

1

0.002

8.9

9

Jo

ur

na l

1000C12A7:e-

9.6

e-

800C12A7:e-

Mean pore diameter (nm)

14.0

Pr

700C12A7:e-

Pore volume (cm3 g-1)

oo f

Sample

57

Table 2. Loading, average crystalline size, dispersion of Ru and apparent activation energy as well as TOF at 225 oC of CO2 methanation over the Ru catalysts supported on C12A7:OH-1, -Al2O3, or C12A7:eDRu (%)c

CO/Ru (%)d

1.76%Ru/400C12A7:OH-1

1.32

63.8

71.1

0.93%Ru/500C12A7:OH-1

0.72

93.6

99.4

1.76%Ru/500C12A7:OH-1

1.29

67.1

2.85%Ru/500C12A7:OH-1

2.51

33.9

3.54%Ru/500C12A7:OH-1

3.33

1.79%Ru/600C12A7:OH-1

1.46

0.42

108.6

0.79

89.2

1.07

34.5

75.3

3.19

e-

76.2

23.2

70.9

6.04

60.0

60.8

93.6

0.84

1.50

60.4

57.3

93.6

0.78

1.52

58.1

53.6

95.3

0.60

1.72%Ru/900C12A7:OH-1

1.60

57.5

50.2

97.5

0.23

1.90%Ru/-Al2O3

1.18

71.2

82.3

105.8

0.39

1.88%Ru/700C12A7:e-

1.92

45.6

38.9

99.8

0.88

2.09%Ru/800C12A7:e-

2.72

35.1

29.5

139.9

0.078

2.06%Ru/900C12A7:e-

2.95

29.8

25.1

145.3

0.052

1.97%Ru/1000C12A7:e-

4.38

23.6

18.8

94.4

2.41

Jo

ur

na l

1.82%Ru/800C12A7:OH-1

a Loading

96.1

27.0

Pr

1.77%Ru/700C12A7:OH-1

Ea TOF at 225 ℃ -1 e (kJ mol ) (10-2 s-1)f

oo f

SRu (nm)b

pr

Catalysta

of Ru measured by ICP is labeled in the name of the catalyst. b Size of Ru particles calculated

from TEM images (500 particles).

c

Dispersion of Ru calculated from average particle size [38].

d

Calculated from CO-pulse titration. e Apparent activation energy was calculated from the temperatures, 58

at which the CO2 conversions were around 10%. f TOF was defined as the number of CO2 molecules

Jo

ur

na l

Pr

e-

pr

oo f

converted per surface Ru atom measured by CO-pulse titration method per second.

59