TiO2 Catalysts: Role and Influence of

Oct 5, 2015 - Aiming at a detailed understanding of the role of metal–support interactions in the selective methanation of CO in CO2-rich reformate ...
0 downloads 0 Views 1MB Size
Subscriber access provided by UQ Library

Article

Selective CO Methanation on Ru/TiO2 Catalysts: Role and Influence of Metal-Support Interactions Ali M. Abdel-Mageed, Daniel Widmann, Sine E. Olesen, Ib Chorkendorff, Johannes Biskupek, and R. Jürgen Behm ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.5b01520 • Publication Date (Web): 05 Oct 2015 Downloaded from http://pubs.acs.org on October 9, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

Selective CO Methanation on Ru/TiO2 Catalysts:

2

Role and Influence of Metal-Support Interactions

3

Ali M. Abdel-Mageed1,§, D. Widmann1, S. E. Olesen2, I. Chorkendorff2, J. Biskupek3, and R.J.

4

Behm1*

5

1

Institute of Surface Chemistry and Catalysis, Ulm University, D-89069 Ulm, Germany

6

2

Department of Physics, Technical University of Denmark, DK-2800 Lyngby, Denmark

7

3

Central Facility of Electron Microscopy, Ulm University, D-89069 Ulm, Germany

8

KEYWORDS: Selective CO methanation, surface area, selectivity, particle size effects, metal-

9

support interactions, Ru/TiO2

10

11

ABSTRACT: Aiming at a detailed understanding of the role of metal-support interactions in the

12

selective methanation of CO in CO2-rich reformate gases we have investigated the catalytic

13

performance of a set of Ru/TiO2 catalysts with comparable Ru loading, Ru particle size and TiO2

14

phase composition, but very different surface areas (ranging from 20 to 235 m2g-1) in this

15

reaction. The activity for CO methanation, under steady-state conditions, was found to strongly

16

depend on the TiO2 support surface area, increasing first with increasing surface area up to a

17

maximum activity for the Ru/TiO2 catalyst with a surface area of 121 m2 g-1 and then decreasing

18

for an even higher surface area, while the selectivity is mainly determined by the Ru particle

ACS Paragon Plus Environment

1

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 43

1

size, which slightly decreases with increasing support surface area. This goes along with an

2

increase in selectivity for CO methanation, in agreement with a model proposed previously for

3

non-reducible supports. In situ infrared measurements further revealed that also the adsorption

4

properties for these catalysts, as evidenced by the CO adsorption strength, change significantly

5

with increasing catalyst surface area and that strong metal-support interactions cause a partial

6

overgrowth of the Ru nanoparticles for the highest surface area catalyst. The interplay between

7

catalyst surface area and reaction characteristics and the important role of metal-support

8

interactions in the reaction, in addition to particle size effects, will be elucidated and discussed.

9

1. Introduction

10

The selective methanation of CO is a key reaction in the transition phase towards an

11

environmentally more friendly H2 economy, providing a simple way for the removal of CO

12

contaminations from H2-rich feed gases for polymer electrolyte fuel cells (PEFCs) that were

13

generated by steam reforming of fossil fuels. Due to the high sensitivity of the Pt based anode

14

catalysts of PEFCs, CO contaminations have to be reduced to below 10 ppm for this purpose. In

15

particular for decentralized small-scale applications, e.g., domestic operation (house heating) CO

16

methanation represents an attractive alternative to the commonly used preferential CO oxidation

17

(PROX) reaction, since it does not require additional features for controlling and dosing

18

additional reactants such as O2.1-3 Because of the high CO2 contents of typical reformate gases,

19

catalysts for this application have to be 100% selective for CO methanation to avoid intolerable

20

losses of H2 due to CO2 methanation. Supported Ru catalysts, e.g., Ru/Al2O3, Ru/zeolite, or

21

Ru/TiO2, are well known for their high activity for CO methanation, while their activity for CO2

22

methanation is rather low.4 Hence, they are attractive candidates for such kind of applications.5-7

23

In addition to its relevance for practical applications, the selective CO methanation represents

ACS Paragon Plus Environment

2

Page 3 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

also an ideal case (model reaction) for evaluating the catalytic performance of metal oxide

2

supported Ru catalysts in hydrogenation reactions in general.8,9

3

Previous studies have demonstrated that both activity and in particular the selectivity of Ru

4

nanoparticle catalysts can be varied in a wide range by varying the Ru particle size, allowing to

5

reach 100% selectivity in CO2-rich reformate gases down to very low CO concentrations.5,7,10-13

6

In the present paper we report results of a different approach, where we tried to modify the

7

metal-support interaction (MSI) in a very controlled way, namely by systematically increasing

8

the surface area of the support, while essentially maintaining other important properties such as

9

Ru loading or Ru particle size. Because of their well-known tendency of exhibiting pronounced

10

MSIs, we used a set of 4 Ru/TiO2 supported catalysts with widely varying TiO2 surface areas,

11

while keeping the Ru loading and particle size as well as the phase composition of the support

12

largely unchanged. In this work we focus on the state and performance of the catalysts under

13

steady-state reaction conditions.

14

Before moving to the results we will briefly summarize the main results of previous studies

15

pertinent for this work. There seems to be general agreement in the meantime that under

16

conditions of low CO contents in the reaction gas, where the selectivity is not determined by site

17

blocking effects, due to blocking of sites for CO2 dissociation by adsorbed CO, the Ru particle

18

size is the decisive parameter for the selectivity of these catalysts.5,7,11 Lower CO2 methanation

19

activities originate from a decreasing CO2 dissociation activity with decreasing Ru particle size,

20

which in turn results in an increase of the selectivity for CO methanation.1,6,11,14-16 This behavior

21

was explained by a lower stability of the CO2 dissociation product COad with decreasing particle

22

size, which according to the Brønsted–Evans–Polanyi (BEP) principle results in a higher

23

activation barrier for CO2 dissociation.11,12,17 The intrinsic activity, in terms of turnover

ACS Paragon Plus Environment

3

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 43

1

frequencies (TOFs), of Ru catalysts for CO methanation, in contrast, decreases with decreasing

2

Ru particle size.7,11,18-20 This was explained by a structural effect, assuming a higher CO

3

methanation activity on the crystallite facets, whose relative abundance decreases with

4

decreasing Ru particle size. For the Ru mass normalized activity, however, this may be different,

5

since a decreasing Ru particle size always goes along with an increase in dispersion. Hence,

6

there is a counterbalance between activity gain (due to higher Ru surface area) and activity loss

7

(due to lower TOFs) with decreasing particle size, which can result in either a decrease or an

8

increase in Ru mass normalized. Both situations have been observed, depending on the exact

9

nature of the catalyst and the conditions during catalyst activation as well as during

10

reaction.11,12,21,22 It was, for example, reported that the increase of Ru particle size resulting from

11

the increase of Ru loading on a Ru/TiO2 and Ru/zeolite catalysts results in an increased Ru mass

12

normalized activity.7,11 On the other hand, formation of larger Ru particles by reduction

13

calcination7 at elevated temperatures resulted in lower Ru mass normalized activities. Wherever

14

TOFs were reported, however, these increased with increasing particle size.

15

It is well known that also the nature of the support material has a pronounced influence on the

16

catalytic performance of supported Ru catalysts in the selective CO methanation. Comparison of

17

differently supported Ru catalysts revealed that Ru catalysts supported on TiO2 are among the

18

most active and selective catalysts for this reaction.3,7,22 Similar findings were reported for other

19

metals (Rh and Pd), where TiO2 supported catalysts also showed significantly higher CO

20

methanation / hydrogenation activities compared to catalysts based on other support materials,

21

e.g., Al2O3, MgO, or SiO2.24-28

22

Already in the late 70s of the last century, Tauster et al. demonstrated that for group VIII small

23

metal particles supported on reducible metal oxides, in particular on TiO2, that the oxide not only

22,23

or

ACS Paragon Plus Environment

4

Page 5 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

acts as an otherwise inert stabilizer for small metal nanoparticles but that interactions between

2

metal und support (strong metal-support interactions (SMSI)), may alter the (electronic and / or

3

structural) properties of the metal nanoparticles upon reduction at elevated temperatures.29-31

4

This may result in drastic changes of their chemisorption properties and catalytic

5

performance.25,30 Therefore, a mechanistic understanding of Ru/TiO2 catalysts and of the

6

ongoing surface processes requires the separation and disentangling of Ru particle size effects

7

and possible metal-support interactions.32

8

In the following we will, after a short description of the experimental procedures (Section 2) and

9

the catalyst characterization (section 3.1), first present results of kinetic measurements,

10

comparing the activity and selectivity of the different Ru/TiO2 catalysts for selective CO

11

methanation in CO2-rich reformates containing medium (6000 ppm) and low (100 ppm) CO

12

concentrations (section 3.2). Next we present results of in situ IR spectroscopy measurements

13

(Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS)), where we determined

14

the (relative) amounts and types of adsorbed surface species formed during the reaction, focusing

15

on the evolution of adsorbed CO species (section 3.3). Finally we evaluated the tendency for CO

16

formation via CO2 dissociation in CO-free CO2-rich reformate gas. Based on these data we will

17

try to disentangle in a systematic way the contributions arising from different effects, including

18

Ru particle size effects, metal-support interactions and possible other effects on the reaction

19

characteristics.

20

2. Experimental

21

2.1 Catalyst preparation and pre-treatment

22

The Ru/TiO2 catalysts were prepared by incipient wetness impregnation using commercially

23

available TiO2 supports with different surface areas (TiO2-1: Sigma Aldrich; TiO2-2: P25 /

ACS Paragon Plus Environment

5

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 43

1

Degussa; TiO2-3: P90 / Degussa; TiO2-4: Sachtleben). For impregnation with Ru, 63.5 mg of the

2

Ru precursor (RuCl3 hydrate, Sigma Aldrich) were dissolved in the appropriate amount of

3

millipore water before adding 1 g of the support material (as received) under stirring. The raw

4

Ru/TiO2 catalysts obtained that way were finally filtrated and dried at 25 °C overnight.

5

Prior to all experiments, the catalysts were pre-treated in-situ by calcination and subsequent

6

reductive activation. Calcination involved first heating up in N2 to 150°C, then exposure for 30

7

min to 10% O2/N2 at 150 °C and finally flushing for 15 min with pure N2 at the same

8

temperature. For the reductive activation, the Ru/TiO2 catalysts were exposed to the respective

9

reaction gas atmosphere during heating up from 150 °C to the reaction temperature (190 °C),

10

after that the reaction was run for over 1000 min.

11

2.2 Catalyst characterization

12

Transmission electron microscopy (TEM)

13

The Ru particle size was determined from TEM imaging performed after reaction in SR-ref 6000

14

reformate at 190 °C for 1000 min on stream, where no more changes in catalytic activity were

15

observed and a steady-state situation is reached (conditions see below). The particles were

16

deposited by drop coating onto holy carbon grids. The TEM measurements were performed on a

17

Philips CM 20 instrument operated at 200 keV. For detailed information on the Ru particle size

18

(volume-area mean diameter and size distribution), at least 500 particles were evaluated for each

19

sample. Moreover, the Ru dispersion was also calculated from the volume-area mean diameter of

20

the Ru nanoparticles. The corresponding equations are given in the Supporting Information (SI).

21

X-ray diffraction (XRD)

22

XRD measurements on the Ru/TiO2 catalysts were performed on a Siemens D5000 spectrometer,

23

using Cu Kα radiation (λ = 0.154 nm). The average TiO2 crystallite sizes of the different catalysts

ACS Paragon Plus Environment

6

Page 7 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

were calculated from the TiO2 (25.373°) diffraction peak of anatase, via the Scherrer equation (D

2

= Κ λ / FWHM cos θ). Here K = 0.89 is the Scherrer’s constant, λ the wave length of the X-rays

3

and FWHM the full width at half maximum of the diffraction peak. The ratio of anatase / rutile

4

was also estimated from the diffractograms, taking the relative intensities of the diffraction peaks

5

at 25.373° (TiO2(101), anatase) and 27.522 (TiO2(110), rutile) as measure. This enables to detect

6

differences in the ratio of the different phases, between the different catalysts.23

7

X-ray photoelectron spectroscopy (XPS)

8

X-ray photoelectron (XP) spectra were recorded on a PHI 5800 ESCA system (Physical

9

Electronics) using monochromatic Al-Kα radiation (1486 eV). Spectra were recorded for the raw

10

catalyst as well as for the samples after reaction (1000 min in SR-ref 6000 at 190 °C).

11

The binding energies (BEs) of all spectra were calibrated against the C(1s) peak, which was

12

fixed at a binding energy of 284.4 eV. The Ru and Ti surface concentrations and their ratio were

13

calculated from the measured intensities of the Ru(3d) and Ti(3p) signals, using tabulated

14

sensitivity factors and assuming a constant composition of the topmost few layers.

15

2.3 Kinetic measurements

16

Kinetic measurements were performed in a quartz tube micro reactor at atmospheric pressure in

17

H2-rich reformate gas mixtures with CO concentrations of either 0.6% (SR-ref 6000: 0.6% CO,

18

3.0% N2, 15.5% CO2, balance H2) or of 0.01% (SR-ref 100: 0.01% CO, 3.0% N2, 15.5% CO2,

19

balance H2). The gas mixtures were prepared using mass flow controllers from Hastings (HFC-

20

202) with a total gas flow of 41.6 Nml min-1 for all measurements. In order to ensure differential

21

reaction conditions the catalysts were diluted with varying amounts of catalytically inactive and

22

thermally stable α-Al2O3 powder (calcined at 1200°C for 24 hr). Depending on the catalytic

23

activity of the respective catalyst and on the reaction atmosphere (SR-ref 6000 or SR-ref 100),

ACS Paragon Plus Environment

7

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 43

1

dilutions of the catalyst with Al2O3 ranged from 1:5 to 1:400, resulting in total CO conversions

2

well below 20% in all measurements. In total, about 200 mg of diluted catalyst was used,

3

resulting in a catalyst bed of ~1.2 cm length.

4

After calcination of the catalyst at 150 °C (see above) the catalysts activity was determined

5

during 1000 min reaction at 190 °C. Here the partial pressures of influent and effluent gases

6

were analyzed by on-line gas chromatography with a CO detection limit of ca. 5 ppm (DANI

7

86.10), using H2 as carrier gas and thermal conductivity detectors.

8

The Ru mass-normalized reaction rate for the CO methanation (CO + 2H2 → CH4 + H2O) was

9

calculated from the CO conversion (XCO) under differential reaction conditions (XCO < 20%, see

10

above), the molar flow rate of CO into the reactor (ṅCO,in), and the absolute mass of Ru metal

11

(mRu) according to equation (1). The Ru mass-normalized CH4 formation rate, in contrast, was

12

calculated by the effluent molar flow rate of CH4 formed (ṅCH4,out), produced from CO and CO2

13

(equation 2). From these Ru mass-normalized reaction rates turnover frequencies (TOFs) were

14

calculated using the molar mass of Ru (MRu) and the Ru dispersion (DRu) obtained from TEM

15

imaging, according to equation 3. The selectivity towards CO methanation (SCO) is given by the

16

ratio of the CO methanation compared to the overall methane formation (from CO and CO2, see

17

equation 4).  × , 

18

 =

19

 =

20

 =

21

 =



,

(2)

  ×

 



=

(1)

  

(3)

(4)

ACS Paragon Plus Environment

8

Page 9 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

2.4 In situ IR spectroscopy measurements

2

In situ diffuse reflectance FTIR spectroscopy (DRIFTS) measurements were performed in a

3

commercial reaction cell (Harricks HV-DR2). The spectra were recorded with a Magna 6700

4

spectrometer (Thermo) equipped with a MCT narrow-band detector. Gas mixtures were prepared

5

as described above for the kinetic measurements, and similar gas flows were used. Spectra were

6

collected in SR-ref 6000 gas reformate. The reaction was followed over 800 min with spectra

7

recorded in situ (400 scans collected per one spectrum). The intensities were evaluated in

8

Kubelka-Munk units which are linearly related to the adsorbate coverage.33 Background removal

9

of the spectra was performed by dividing the measured absorption by the absorption obtained in

10

spectra recorded in a flow of N2 at 150°C, before heating up to 190 °C in reaction atmosphere.

11

3. Results and discussion

12

3.1 Ru/TiO2 catalyst characterization

13

The Ru loadings of the various Ru/TiO2 catalysts used in this study, as determined by inductively

14

coupled plasma optical emission spectroscopy (ICP-OES), are 2.8, 2.2, 2.2, and 2.1 wt.% for

15

Ru/TiO2-1, Ru/TiO2-2 and Ru/TiO2-3, and Ru/TiO2-4, respectively (see Table 1) and hence

16

rather similar to each other. The surface areas of the Ru catalysts, in contrast, differ significantly,

17

ranging from 20 m2 g-1 to 235 m2 g-1 for Ru/TiO2-1 and Ru/TiO2-4, respectively. Accordingly,

18

also the TiO2 crystallite sizes of the samples differ significantly from each other. Based on XRD

19

measurements the crystallite size of the TiO2 particles decreases from 52 nm for Ru/TiO2-1 to

20

about 9 nm for Ru/TiO2-4, respectively. This trend fits quite well, at least qualitatively, to the

21

observed increase in surface area. The XRD measurements provide additional information on the

22

phase composition of the support material. For all support materials the most prominent

ACS Paragon Plus Environment

9

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 43

1

diffraction peaks were detected at 25.3° and 48.1°, which are related to the (101) and (200)

2

planes of anatase.34,35 Hence, in all samples TiO2 is predominantly present as anatase. In

3

addition, there are also weaker reflections at 27.4°, 36° and 55°, which are assigned to the (110),

4

(101) and (211) planes of rutile. The diffractograms indicate that in all catalysts TiO2 is present

5

as a mixture of anatase and rutile, as it is well known for TiO2-2 (P25). Taking the intensity ratio

6

of the (101) reflection of anatase and of the (110) reflection of rutile as a measure, there are no

7

significant differences in the phase composition of the four Ru/TiO2 catalysts (see Table 1 and

8

Figure S1, Supporting Information).

9

The chemical composition of the catalyst surfaces was characterized by XPS measurements,

10

evaluating the Ru(3d) and Ti(2p) signals of the four samples after calcination and reaction for

11

1000 min in SR-ref 6000 reformate gas at 190 °C. The spectra presented in Figure S2 show that

12

the binding energy (BE) for the 3d5/2 state of Ru is between 279.5 and 280.0 eV, which refers to

13

a largely reduced catalyst.36 The BEs of the Ti (2p3/2) signals are between 458.1 eV and 458.8

14

eV, which is typical for Ti4+.37,38 The relative ratio of oxidic to reduced Ru species as well as the

15

expected shifts in the core shell could not be determined by XPS because of instantaneous

16

surface oxidation upon exposure to air, during catalyst transport to the spectrometer. In-situ

17

XANES spectra on Ru/TiO2-3 at the Ru K-edge, which were recorded under reaction conditions,

18

however, revealed mainly metallic Ru species during CO methanation. Based on these results we

19

consider Ru to be present as metallic species under reaction conditions, as it was previously

20

reported also for Ru/Al2O3 and Ru/zeolite catalysts using XANES / EXAFS experiments.12,39

21

ACS Paragon Plus Environment

10

Page 11 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

Table 1. Physical properties, catalytic activities for the selective CO methanation at 190 °C (in

2

SR-ref 6000) and selectivities for CO methanation (in SR-ref 100) of various Ru/TiO2 catalysts

3

with different specific surface areas.

Parameter Ru loading / wt. % Surface area / m2 g-1 Pore diameter / nm

Ru/TiO2-1

Ru/TiO2-2

Ru/TiO2-3

Ru/TiO2-4

2.8

2.2

2.2

2.1

20

64

121

235

3.45

3.08

3.05

2.27

Cl content / % (XPS) Ru / Ti (atomic ratio / XPS) TiO2 crystallite size / nm

0.32

0.47

0.34

0.50

0.076

0.043

0.016

0.012

52.4

19.3

14.0

8.8

Ratio anatase / %

92

83

89

97

Ru particle size / nm

2.3

1.8

1.6

1.5

Ru dispersion / %

43

55

68

71

Reaction rate / 10-6 mol gRu-1 s-1

2.7

20

30.0

11.1

TOF / 10-3 s-1

0.65

3.7

4.4

1.8

Selectivity / %

58

76

90

95

4

Highly important is also the atomic ratio of Ru : Ti atoms in the topmost layers of the Ru/TiO2

5

catalysts, which was calculated from the intensity ratio of the Ru (3d5/2) and Ti (2p3/2) peaks,

6

taking into account the tabulated sensitivity factors for the respective signals. The Ru : Ti ratio is

7

found to decrease continuously with increasing surface area, by a factor of about 3 when going

8

from Ru/TiO2-1 to Ru/TiO2-4 (see Table 1). This significant decrease of the Ru : Ti ratio at

ACS Paragon Plus Environment

11

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 43

1

(almost) constant Ru loading and Ru dispersion is clearly correlated with the increase of support

2

surface area.

3 4

Figure 1. TEM images and particle size distributions of Ru NPs supported on TiO2 with different

5

surface areas: a) Ru/TiO2-1 b) Ru/TiO2-2; c) Ru/TiO2-3, and d) Ru/TiO2-4. Measurements were

6

performed ex-situ, after reaction in SR-ref 6000 gas reformate at 190°C for 1000 min (Ru/TiO2-

7

1: black; Ru/TiO2-2: red; Ru/TiO2-3: green; Ru/TiO2-4: blue).

8

TEM images taken of all samples after reaction in SR-ref 6000 for 1000 min at 190°C are shown

9

in Figure 1, together with the corresponding particle size distributions. These images clearly

10

illustrate that the TiO2 nanoparticles decrease in size with increasing surface area, in agreement

11

with the XRD results, and that the Ru nanoparticles are well separated and highly dispersed on

ACS Paragon Plus Environment

12

Page 13 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

the TiO2 surface. The Ru size distributions reveal that the mean Ru particle size decreases

2

slightly with increasing surface area, from 2.3 nm (Ru/TiO2-1) to 1.5 nm (Ru/TiO2-4). While the

3

differences in Ru particle size are rather small for Ru/TiO2-2, Ru/TiO2-3 and Ru/TiO2-4 (1.8,

4

1.6, and 1.5 nm, respectively), the mean Ru particle size for Ru/TiO2-1 catalyst (2.3 nm) differs

5

significantly (see Table 1). Based on the particle size distributions we also calculated the

6

dispersions, assuming hemispherical Ru particles and 1.5×1015 Ru atoms cm-2. The resulting Ru

7

dispersions are 43, 55, 68, and 71% for the Ru/TiO2-1, Ru/TiO2-2, Ru/TiO2-3 and Ru/TiO2-4

8

catalysts, respectively. Here one should note that the Ru surface area may be slightly

9

underestimated due to partial overgrowth by TiOx caused by strong-metal support interactions.

10

This is in particular true for the catalyst with the highest surface area / smallest TiO2 particles

11

(Ru/TiO2-4), which will be discussed below (section 3.3).

12

Overall, the differences in Ru loading, Ru particle size (or Ru dispersion) and TiO2 phase

13

composition of the four Ru/TiO2 catalysts are rather small compared to the differences in catalyst

14

surface area, which grows by more than one order of magnitude from Ru/TiO2-1 to Ru/TiO2-4,

15

from 20 to 235 m2 gcat-1. The influence of the different structural properties on the reaction

16

characteristics will be discussed in more detail below, after presenting the results of the kinetic

17

and in situ IR spectroscopic studies.

18

3.2 Activity and selectivity for CO methanation

19

The performance of the Ru/TiO2 catalysts in the selective methanation reaction was determined

20

in kinetic measurements at atmospheric pressure and at 190°C in two different reaction

21

atmospheres. First the catalytic activity for CO methanation was determined in SR-ref 6000

22

(6000 ppm, 3.0% N2, 15.5% CO2, balance H2). In this reformate, high selectivities of essentially

ACS Paragon Plus Environment

13

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 43

1

100% are known to result from surface blocking by adsorbed CO, which blocks the active sites

2

for CO2 adsorption and dissociation, as had been demonstrated in previous measurements on

3

Ru/Al2O3 and Ru/zeolite catalysts.6,12 Similar results were obtained for the Ru/TiO2 catalysts

4

investigated in the present study in SR-ref 6000, and they are similarly attributed to site blocking

5

by COad. Therefore, the inherent selectivity of the catalyst for CO methanation in the presence of

6

large amounts of CO2 (15.5%) has to be determined in a reformate containing much lower CO

7

contents (SR-ref 100: 100 ppm CO, 3.0% N2, 15.5% CO2, balance H2). This low CO partial

8

pressure results in a much lower COad coverage on the Ru surface. Accordingly, there are active

9

sites available for CO2 dissociation and CO2 methanation, and the selectivity is no longer

10

governed by adlayer site blocking. Technically, such low CO concentrations are met at the end

11

of methanation reactor, and also under these condition a high selectivity is mandatory to avoid

12

intolerable losses of H2 due to methanation of CO2.6

13

First the activity of the different Ru/TiO2 catalysts was evaluated during 1000 min time on

14

stream in SR-ref 6000, until almost no changes in activity were detected anymore and steady-

15

state conditions were reached. The temporal evolution of the Ru mass normalized activities of

16

the different catalysts is illustrated in Figure S3. The data show for all catalysts an initial

17

activation period, ranging over the first 50 - 300 min of the reaction, where the activity increases

18

with time on stream. This is followed by a continuous decrease in reaction rate (deactivation).

19

This temporal evolution (activation / deactivation) and its origin are topic of ongoing work in our

20

laboratory, whose results will be reported later. In the present paper, we focus on the catalytic

21

activities of the different supported Ru catalysts under steady-state conditions, after 1000 min

22

reaction at 190 °C (see Figure 2b). For the first three catalysts the Ru mass normalized activity

23

increases continuously with increasing support surface area, from 2.7×10-6 mol gRu-1 s-1 for the

ACS Paragon Plus Environment

14

Page 15 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

20 m2 g-1 (Ru/TiO2-1) catalyst to 30.0×10-6 mol gRu-1 s-1 for the 121 m2 g-1 (Ru/TiO2-3), catalyst

2

(see Table 1). A further increase of the catalyst surface area to 235 m2 g-1, however, results in an

3

unexpected loss of activity to only 11.1×10-6 mol gRu-1 s-1 (Ru/TiO2-4). Overall this results in a

4

volcano type dependence of the CO methanation activity on the catalyst surface area, with the

5

highest reaction rate observed on the Ru/TiO2-3 catalyst with a surface area of 121 m2 g-1. Note

6

that for the latter catalyst the Ru mass normalized activity is comparable to that reported for a

7

commercial, highly active and selective Ru/zeolite catalyst (Clariant Produkte).11,12

8

We also calculated the TOFs of the various catalysts, using the Ru dispersion as calculated from

9

the Ru particle size distribution (see section 3.1) in order to determine the intrinsic catalytic

10

activities of the Ru nanoparticles. As shown in Figure 2, these exhibit the same volcano like

11

dependence on the catalyst surface area as the Ru mass normalized activities (see Table 1).

12

While the TOF increases by a factor of about 7, from 0.65×10-3 s-1 to 4.4×10-3 s-1, when going

13

from the Ru/TiO2-1 to the Ru/TiO2-3 catalyst, it is again much lower, by a factor of about 2.5,

14

for Ru/TiO2-4 (1.8×10-3 s-1). Even when considering an underestimation of the dispersion for

15

Ru/TiO2-4 by 20% (due to the partial overgrowth of the Ru NPs by reduced TiOx, see also

16

section 3.3), the TOF of the latter catalyst is still significantly lower compared to Ru/TiO2-2 and

17

Ru/TiO2-3.

18

One possible explanation for such a dependency of TOFs on the Ru particle size would be that

19

special undercoordinated sites, e.g., step or edge sites, on the Ru nanoparticles are the active sites

20

for CO methanation. Such behavior was demonstrated, e.g., for the ammonia synthesis on

21

supported Ru catalysts, and special B5-type sites were held primarily responsible for their

22

catalytic activity.40 Here the relative amount first increases with decreasing Ru particle size,

ACS Paragon Plus Environment

15

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 43

1

down to about 2 nm, but decreases again for even smaller Ru particles.40,41 The resulting

2

structure sensitivity in the Ru catalyzed ammonia synthesis would fit to the Ru size dependency

3

of the (TOF based) activity reported above. Similarly, also for the CO methanation on Ni

4

catalysts undercoordinated sites were proposed to represent the active sites for the rate limiting

5

step (dissociation of a COH intermediate).42 On the other hand, previous studies on the Ru

6

catalyzed CO methanation found that the TOF based activity increases with increasing Ru

7

particle size,5,7,11,12,17,19,43 which was explained by the higher activity of planar facets.

ACS Paragon Plus Environment

16

80

2.2

70

2.0

60

1.8

50

1.6

32

4

24

3 2

16

TOF / 10-3 s-1

b)

1

8

0

0 36 c)

100

30

90

24

80

5 70

4 3

60

CO selectivity/ %

-1 -1

mol g Ru s

-6

Rate / 10 -1

40

-1

mol g Ru s

-6

Ru dispersion / %

2.4 a)

1.4

Rate / 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Ru particle size / nm

Page 17 of 43

2 50 50

1

100

150

200

SA / m2 g-1

2

Figure 2. a) Average Ru particle diameter (▼) and Ru dispersion (○) after reaction for 1000 min

3

in SR-ref 6000 at 190°C, b) steady state Ru mass normalized CH4formation rates (●) in SR-ref

4

6000 gas mixture at 190°C and corresponding turn over frequencies (♦), c) Ru mass normalized

5

activity (CO conversion: ■; CH4 formation: ●) and corresponding CO selectivity (▲) in SR-ref

6

100 reformate at 190°C, all as a function of TiO2 surface area (SA).

ACS Paragon Plus Environment

17

ACS Catalysis

1

The present finding of a maximum activity at a particle size of 1.6 nm and a steep decay of the

2

activity for a catalyst with only slightly smaller particles (Ru/TiO2-4) seems to be specific for

3

Ru/TiO2 catalysts, and correlated also with the significant variation in TiO2 surface area. Hence,

4

we assume that the distinct increase in activity (mass normalized activity as well as TOF) with

5

increasing surface area and decreasing particle size (Ru/TiO2-1, Ru/TiO2-2, and Ru/TiO2-3),

6

does not originate from particle size effects alone, but includes also contributions from the

7

support surface area. The latter will be discussed in more detail below, after evaluation of the

8

adsorption properties of the different catalysts. Finally we would like to note that differences in

9

the TiO2 phase composition can also be excluded as reason for the observed variations in the

10

catalytic performance, since this is rather similar for all catalysts, and does not follow at all the

11

trend connected in the catalytic activity.

100 CO Conversion / %

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 43

80 60 40 20 0

12

200 220 240 260 Reaction Temperature / °C

13

Figure 3. Relative CO conversion during a temperature programmed reaction (190 °C – 270 °C)

14

in SR-ref 6000 reformate gas of Ru/TiO2-1 (●), Ru/TiO2-2 (▲), Ru/TiO2-3 (◊), and Ru/TiO2-4

15

(♦) after reaching a steady –state situation at 190 °C.

ACS Paragon Plus Environment

18

Page 19 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

In the above measurements, all catalysts were 100% selective for CO methanation due to COad

2

site blocking (see above in this section).

3

To gain further information on temperature effects, we performed another set of experiments,

4

where the reaction temperature was increased from 190 to 270°C during the CO methanation in

5

SR-ref 6000, after having reached a steady-state situation at 190°C (temperature programmed

6

reaction, heating ramp 5°C min-1). While for the first three catalysts the CO conversion increased

7

continuously with temperature over the whole temperature range, until reaching full conversion,

8

the Ru/TiO2-4 with the highest surface area (235 m2 g-1). catalyst showed a moderate increase in

9

CO conversion only from 190°C to 220°C, followed by a decrease in activity for even higher

10

temperatures (see Figure 3). At that temperature, the catalyst becomes increasingly active for the

11

reverse water gas shift reaction, producing mainly CO and H2O and no longer CH4. The other

12

three catalysts, in contrast, were only active for the methanation reaction (CO and CO2

13

methanation) in the entire temperature range investigated. Up to at least 220°C, all catalysts are

14

100% selective for CO methanation. For higher temperatures, CO2 methanation sets in, most

15

likely due to a decreasing COad coverage, with different relative contributions for the different

16

catalysts.

17

The activities of the four Ru/TiO2 catalysts for CO methanation as well as for total methane

18

formation, from CO and CO2, in SR-ref 100 at 190°C are shown in Figure 2c, together with the

19

corresponding selectivities. The trend in the total methane formation activity is similar to that

20

described above for reaction in SR-ref 6000. The activity first increases from Ru/TiO2-1 to

21

Ru/TiO2-3, reaching a maximum activity for the latter catalyst, and then decreases again for the

22

Ru/TiO2-4 catalyst. This decrease in activity, however, is much less pronounced compared to

23

reaction in SR-ref 6000. Another remarkable difference is the activity for methane formation

ACS Paragon Plus Environment

19

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 43

1

from CO2, which is given by the difference between CO conversion and methane formation. This

2

difference is most prominent for the Ru/TiO2-1 and Ru/TiO2-2 catalysts (see Figure 2c), and

3

hardly present any longer for Ru/TiO2-4. Calculations of the respective selectivities for CO

4

methanation show that these are 58, 76, 90, and 95% on Ru/TiO2-1, Ru/TiO2-2, Ru/TiO2-3, and

5

Ru/TiO2-4, respectively. Hence, in contrast to the catalyst activity, there is no maximum for

6

Ru/TiO2-3, but a steady increase with increasing surface area of the catalysts. Taking into

7

account previous findings on the origin of the selectivity, as described in the introduction, this

8

increase in selectivity is proposed to originate at least partly from the decreasing Ru particle size,

9

which decreases in the same order, from 2.3 nm (Ru/TiO2-1) to 1.5 nm (Ru/TiO2-4). Although

10

these changes in Ru particle size are rather small, it is well known from previous studies on other

11

Ru catalysts that such changes may be sufficient to induce significant changes in

12

selectivity.5,7,11,12,44 Similar results of an increasing selectivity for CO methanation with

13

decreasing Ru particle size have been reported in various previous studies, where variation of the

14

Ru particle size was realized by i) a decreasing Ru loading on Ru/zeolite,11 ii) an increasing

15

calcination temperature on Ru/zeolite, 12 iii) a decreasing reduction temperature on Ru/Al2O3 and

16

Ru/TiO2,3,7 or iv) a pre-treatment in the presence of water in the reaction feed on Ru/Al2O3.17

17

Hence, the present observations are in good agreement with previous findings and clearly

18

demonstrate a general dependence of the CO selectivity on the Ru particle size, independent of

19

the nature of the metal oxide support (reducible or non-reducible) or exact procedures for

20

catalyst preparation and pre-treatment. This behavior was previously explained by a lower

21

activity of small Ru nanoparticles for the dissociation of CO2, which was attributed to a lower

22

stability of the dissociation products of CO2 (COad + Oad) on these very small particles.11,12,17

ACS Paragon Plus Environment

20

Page 21 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

According to the Brønstedt-Evans-Polanyi(BEP) relation the activation barrier for the CO2

2

dissociation increases with decreasing stabilization of the final product COad + Oad.45

3

Changes in the selectivity for CO in the presence of large amounts of CO2 have also been

4

reported by Miyao et al. and 46 and Djinovic et al. 47,48 for Ru/Al2O3, Ru/CeO2 and Ni-Al mixed

5

oxide catalysts, respectively. They demonstrated that the presence of significant amounts of

6

(residual or doped) Cl increases the selectivity, which they explained by Clad site blocking of the

7

metal active sites for CO2 dissociation to COad 46. Accordingly, one may speculate that also for

8

the present Ru/TiO2 catalysts residual Cl species on the catalyst surface, which might be present

9

after catalyst preparation and activation, influence their catalytic performance, which would lead

10

to the observed differences. XPS measurements performed on the four Ru/TiO2 catalysts after

11

activation revealed, however, rather similar Cl surface concentrations of 0.3 - 0.5% for all

12

samples (see Table 1). No Cl at all was detected in EDX measurements. Hence, considering the

13

similar amount of Cl and the absence of any correlation between the Cl content and the

14

selectivity, a significant impact of residual Cl can essentially be ruled in the present study.

15

3.3 Adlayer formation

16

In order to gain additional information on the chemical properties of the Ru nanoparticles and

17

possible changes therein with increasing surface area of the catalysts, we followed the build-up

18

of adsorbed species during the reaction by in-situ IR spectroscopy. In situ DRIFT spectra were

19

recorded upon interaction of the different Ru/TiO2 catalyst samples with the SR-ref 6000

20

reformate at 190°C until reaching steady-state conditions (after 800 min), applying identical pre-

21

treatment and reaction conditions as described above for the kinetic measurements (see section

22

3.2). Since we were particularly interested in the nature and the amount of adsorbed CO, formyl

ACS Paragon Plus Environment

21

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 43

1

and formate species formed on the catalyst surface during reaction, we focus here on the spectral

2

region between 1200 cm-1 and 2200 cm-1. The evolution of these species, which are known to act

3

as reaction intermediate (COad, formyl species) and spectator species (formate), respectively,39,49

4

can provide valuable information on the differences in the catalytic performance. Sets of spectra

5

showing this region, which were recorded during the CO methanation at different reaction times

6

(from 0 to 735 min) on the different catalysts, are presented in Figure 4. Sets of spectra covering

7

the OH-region (around 3700 cm-1) and the CHx-region (around 3000 cm-1) are presented in the

8

Supporting Information (Figure S4).

9

From the spectra presented in Figure 4 it is obvious that both the type as well as the amount of

10

surface species formed on the catalyst surface during reaction differ significantly for the different

11

catalysts. A weak band appearing in the range 2135 -2145 cm-1, together with another band at

12

around 2078 cm-1, is considered to be characteristic for the CO vibration in Ru-multicarbonyl

13

surface species.50,51

ACS Paragon Plus Environment

22

Page 23 of 43

a) 2084

2019

2052 0.1 KMU b)2078 2009

1928

2145

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

c)

2078

2135

1766

2060 2009 1978

0.2 KMU 1556

2000

d) 2075 2058 2009 1982

1764

1800

0.2 KMU 1558

2140

2138

1600

2000 Wavenumber / cm -1

0.2 KMU 1556

1764

1800

1600

1 2

Figure 4. DRIFT spectra recorded during the selective CO methanation (SR-ref 6000 gas

3

mixture, 190°C) on the different Ru/TiO2 catalysts after different reaction times (from bottom to

4

top after 0, 1, 5, 9, 15, 75, 135, 255, 495, 615, and 735 min, respectively. a) Ru/TiO2-1, b)

5

Ru/TiO2-2, c) Ru/TiO2-3, and d) Ru/TiO2-4.

6

For all samples, the intensity of these two bands was found to reach a maximum within the first

7

minutes of reaction, followed by a continuous decrease with time on stream afterwards.

8

Moreover, this band is much more intense on the catalysts with higher surface areas (Ru/TiO2-3

9

and Ru/TiO2-4) than on the samples with lower surface areas. When reaching steady-state

10

conditions, this band has completely vanished for the Ru/TiO2-1 catalyst. In the range from 2060

11

to 1980 cm-1 there are various other surface species which are attributed to CO adsorbed in an

ACS Paragon Plus Environment

23

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 43

1

on-top configuration on Ru particles. While the band at around 2058 cm-1 was assigned to on-top

2

adsorbed CO on small Ru clusters,52-54 the band in the range of 2009 - 2019 cm-1 was proposed

3

to be characteristic for a COad species adsorbed on slightly larger Ru particles.54,55 The respective

4

intensities of these bands (multicarbonyl and COad), as indicated by the corresponding peak

5

heights (KM units) for all four catalysts, are illustrated in Figure 5.

6

Despite the similar Ru loadings and Ru dispersion a quantitative comparison of the absolute

7

amount of adsorbed surface species between different catalysts is hardly possible from these

8

spectra. On a qualitative scale, all bands for COad are most prominent on the Ru/TiO2-3 catalyst,

9

which shows also the highest activity for CO methanation. On the less active Ru/TiO2-2 and

10

Ru/TiO2-4 catalysts, in contrast, the band intensities are much lower. For the sample with the

11

lowest surface area and lowest activity (Ru/TiO2-1) they are hardly visible under steady-state

12

conditions. Moreover, for the Ru/TiO2-3 catalyst, and to a lesser extent also for Ru/TiO2-4, there

13

is also a characteristic peak at 1982 cm-1, which was previously attributed to a COad species

14

adsorbed on-top of highly dispersed, low coordinated Ru clusters isolated from other metal

15

particles.54,56,57 A band characteristic for bridge bonded COad species, located at 1932 cm-1, was

16

only observed on Ru/TiO2-1. This latter finding agrees well with results from a number of

17

previous studies where the relative population of on top and bridge bonded adsorbed CO was

18

reported to depend sensitively on the Ru particle size, with a higher fraction of bridge bonded

19

COad for larger particles.17,36,58,59 Hence, the presence of larger Ru particles increases the

20

possibility of adsorption in a bridge bonded configuration.60,61

ACS Paragon Plus Environment

24

Page 25 of 43

a) 18 12 Intensity / 10-2 KMU

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

6 0

b)

15

10

5

0

0

50

100

150

200

250

SA / m2 g-1

1 2

Figure 5. Intensity of adsorbed CO surface species (peak heights) during reaction in SR-ref 6000

3

after 1000 min (steady state) as a function of the support surface area: a) Ru-carbonyl associated

4

bands (2078 cm-1) and b) on top adsorbed COad (2060 - 2052 cm-1 and 2009 - 2025 cm-1; empty

5

and filled columns, respectively).

6

A band appearing at 1765 cm-1 had been assigned to adsorbed formyl surface species (HCOad)

7

which are formed upon reaction between adsorbed COad and Had.49,62,63 This species had been

8

identified as an active reaction intermediate, since the rate of its off-reaction, as identified in an

9

isotope exchange experiment switching from

12

CO/H2 to 13CO/H2, was rather similar to the CO

10

methanation rate.49 This species appears for all catalysts, but with different intensities depending

11

on the catalyst, as shown in Figure 4. Its coverage first increases gradually with increasing

ACS Paragon Plus Environment

25

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 43

1

support surface area, from Ru/TiO2-1 to Ru/TiO2-3, but decreases again for even higher surface

2

area (Ru/TiO2-4, see Figure S5a). Hence, also the intensity of this species is correlated in the

3

same way with the activity for CO methanation (see Figure 2a) as described above for the COad

4

species.

5

Finally, a band related to surface formates was observed at 1554 cm-1.61,64 The band intensity

6

related to this species increases with increasing surface area (Figure S1b), from being absent on

7

the lowest surface area catalyst (Ru/TiO2-1) to low intensities on the Ru/TiO2-2 and Ru/TiO2-3

8

catalysts to a very pronounced band for the Ru/TiO2-4 catalyst with the highest surface area. For

9

the latter catalyst it is by factors of 15 and 13 higher compared to the Ru/TiO2-2 and Ru/TiO2-3

10

catalysts, respectively. The much higher steady-state coverage of surface formates on the

11

Ru/TiO2-4 catalyst is suspected as a potential reason for the much lower activity of the Ru/TiO2-

12

4 catalyst.

13

In a simple picture the largely different COad band intensities on the different catalysts may be

14

expected to result from significant differences in the COad coverages due to a modification of the

15

CO adsorption properties of the Ru nanoparticles. Considering, however, that for all catalysts

16

CO2 methanation is completely suppressed in SR-ref 6000 due to surface blocking by a dense

17

CO adlayer, resulting in 100% selectivity, the actual COad coverages on the different samples

18

cannot differ to an extent as indicated by the very different IR intensities in Figs. 4 and 5. Most

19

simply, the rather low COad band intensities in particular of the highest surface area (Ru/TiO2-4)

20

catalyst may be due to a significantly lower number of accessible adsorption sites than expected

21

from the dispersion of the Ru particles. This may be caused by a partial overgrowth of the Ru

22

nanoparticles by TiOx species. Such behavior is a characteristic feature of strong metal-support

23

interactions (SMSI), and had been reported earlier, e.g., for oxide supported Pt or Ru

ACS Paragon Plus Environment

26

Page 27 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

catalysts.29,31 Experimental evidence for such SMSI effects in the present case is provided by

2

high resolution TEM images, which are shown in Figure 6. For Ru/TiO2-4, the image resolves

3

lattice fringes of TiO2 crystallites through Ru nanoparticles as also indicated by the low contrast

4

between Ru and TiO2 particles in the TEM micrographs, which is indicative of a partial

5

overgrowth of Ru by small TiOX islands. Similar features are detected also for the Ru/TiO2-3

6

catalyst, but to a much lower extent, while for the other catalysts with even lower surface area

7

(Ru/TiO2-1 and Ru/TiO2-2) they could not be observed. Hence, both the low COad absorption

8

intensity and the TEM images point to a partial overgrowth of the Ru nanoparticles on the

9

Ru/TiO2-4 catalyst, which can explain also the low activity of this catalyst.

10

11

ACS Paragon Plus Environment

27

ACS Catalysis

1

Figure 6. High resolution TEM images of Ru nanoparticles particles on different titania supports

2

(a: Ru/TiO2-1; b: Ru/TiO2-2; c: Ru/TiO2-3; d: Ru/TiO2-4).

3

Further information on the adsorption properties of the various Ru catalysts can be obtained by

4

determining the stability of the CO adlayer under reaction conditions. For this purpose, we

5

performed another experiment where we switched from reaction atmosphere to inert atmosphere

6

(N2 only) at 190 °C. The catalyst surface was first saturated with COad / carbonyl species during

7

reaction for 250 min, before their decomposition / desorption with time was followed for 800

8

min under isothermal conditions.

a)

0.05 KMU b)

0.5 KMU

195 min

15 min

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 43

0 min

0 min

c)

05 KMU

d) 805 min

0.5 KMU

435min

0 min

0 min

2000

1800

2000

1800

Wavenumber / cm -1

9

ACS Paragon Plus Environment

28

Page 29 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

Figure 7. Series of DRIFT spectra recorded during 750 min desorption in N2 at 190°C (0, 1, 3, 5,

2

7, 9, 15, 75, 105,195, 255, 435, 495, 675, and 795 min) after surface saturation in SR-ref 6000 at

3

190°C for 300 min on a) Ru/TiO2-1, b) Ru/TiO2-2, c) Ru/TiO2-3, and d) Ru/TiO2-4).

4

The spectra recorded during desorption in N2, and after saturation of the catalyst surface in SR-

5

ref 6000, are presented in Figure 7. The temporal evolution of the COad bands (see Figure 8)

6

reveals an increasing stability of the COad species adsorbed on the Ru nanoparticles with

7

increasing surface area of the respective catalyst. The time required for complete desorption of

8

the COad species clearly increases with increasing support surface area. While they were almost

9

completely desorbed in less than 10 min for the lowest surface area catalyst (Ru/TiO2-1), it took

10

about 250 and 450 min to reach a COad / Ru-carbonyl free surface for the Ru/TiO2-2 and

11

Ru/TiO2-3 catalysts, respectively. On the Ru/TiO2-4 catalyst, even after 800 min flushing in N2

12

there were still about 50% of the adsorbed CO / Ru-carbonyl species present (see Figure 7 and

13

8).

14

This increasing stability of COad on Ru nanoparticles with increasing surface area of the catalysts

15

(Ru/TiO2-1 < Ru/TiO2-2 < Ru/TiO2-3 < Ru/TiO2-4) cannot be explained by a geometric

16

reduction of the number of accessible surface sites, but points to electronic effects, e.g., by an

17

increasing electron donation to COad species, which will affect both the Ru-CO and the C-O

18

bond. Such a transfer of electronic charge from the support into metal nanoparticles was

19

demonstrated for the first time by Raymond et al. for nickel metal particles supported on

20

germanium. In that case a transfer of only one electron from the support to about 10,000 Ni

21

atoms was proposed to result in an increase of the activity for formic acid decomposition by a

22

factor of 1.3 to 3.0.65 The increase in CO adsorption strength demonstrated above points to such

23

kind of metal-support interactions, which increase with increasing catalyst surface area. Since in

ACS Paragon Plus Environment

29

ACS Catalysis

1

the present work the increase in surface area of the catalysts goes along with a slight decrease in

2

Ru particle size, we obtain an increase of the CO adsorption strength with decreasing Ru particle

3

size, which is opposite to the trend derived previously on other metal oxide supported Ru

4

catalysts (Ru/zeolite,11 Ru /Al2O3 6) Furthermore, the change in electronic and chemical

5

properties due to metal-support interactions may also be responsible for the increasing activity

6

for CO methanation with increasing surface area, at least for surface areas up to 121 m2 gcat-1

7

(Ru/TiO2-1 to Ru/TiO2-3). For the Ru/TiO2-4 catalyst with the highest surface area, however, the

8

measured activity becomes significantly lower because of the partial overgrowth of Ru

9

nanoparticles by TiOx species, as described above.

5 a) Ru/TiO2-1

b) Ru/TiO2-2

4

4

3

2

2

0

60

40 0

3

6

9

1

8

0

0

60 c) Ru/TiO -3 2

d) Ru/TiO2-4 60

40

Intensity / 10-2 KMU

Intensity / 10-2 KMU

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 43

40

20

20

0

0 0

10

200

400

600

800 0

200

400

600

800

Time / min

ACS Paragon Plus Environment

30

Page 31 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

Figure 8. Time evolution of the intensity of adsorbed CO surface species (peak heights) during

2

desorption in pure N2 after switching from SR-ref 6000 as a function of the support surface area

3

(▲: 2140 – 2135 cm-1;●: 2078 cm-1; ○: 2056 - 2050 cm-1;: 2009 - 2025 cm-1; ▼: 1982 cm-1) on

4

a) Ru/TiO2-1, b) Ru/TiO2-2, c) Ru/TiO2-3, and d) Ru/TiO2-4).

5

Here we would like to note that in this case shifts in the binding energies of the Ru nanoparticles

6

(3d and 3p) in the XPS measurements (see section 3.1) are not reliable for identifying electronic

7

modifications related to the changes in MSI indicated by the IR and TEM measurements. This is

8

hindered by the high sensitivity of Ru to oxidation by air during transfer to the XPS chamber.

9

Overall, these results indicate increasing metal-support interactions with increasing TiO2 support

10

surface area / decreasing TiO2 crystallite size, which result in a stabilization of adsorbed CO and

11

which are presumably also responsible for the observed increase in TOF based CO methanation

12

activity. In addition to these electronic modifications, strong-metal support interactions (SMSI)

13

lead to a partial overgrowth of Ru by TiO2, which significantly reduces the number of accessible

14

Ru surface sites and hence the number of adsorbed CO and formyl species as well as the CO

15

methanation activity.

16

3.4 Interaction of Ru/TiO2 with CO2 and origin of selectivity

17

So far we have not addressed the increasing selectivity with increasing surface area of the

18

various Ru/TiO2 catalysts observed for reaction in SR-ref 100, i.e., under reaction conditions

19

where the inherent activity / selectivity of the catalyst is dominant rather than COad site blocking.

20

According to the Brønstedt-Evans-Polanyi (BEP) principle,45 a higher stability of the

21

dissociation product (COad) should result in a lower activation barrier for CO2 dissociation and,

22

hence, in a lower selectivity of the catalyst for CO methanation under these conditions. This,

ACS Paragon Plus Environment

31

ACS Catalysis

1

however, is in contrast to the finding in the kinetic measurements described above, where the

2

selectivity was found to increase in the order Ru/TiO2-1 < Ru/TiO2-2 < Ru/TiO2-3 < Ru/TiO2-4

3

with decreasing Ru particle size, while the CO adsorption strength increased. On the other hand,

4

in previous studies on closely related Ru/zeolite and Ru/Al2O3 catalysts, the selectivity was also

5

observed to increase significantly with decreasing Ru particle size, up to 100% for 1 nm Ru

6

particle size, but in that case the CO adsorption strength decreased with decreasing particle

7

size.11,12 Hence, in that case expectations based on the BEP principle agreed with the

8

experimental finding. To better understand this apparent discrepancy, we directly tested and

9

compared the CO2 dissociation activity of all Ru/TiO2 catalysts by in-situ IR spectroscopic

10

measurements in CO free CO2-ref (15.5% CO2, balance H2).

a)

0.1 KMU

2005

b)

c)

0.2 KMU 2072 1995

2000

0.2 KMU 2075 2001

1868

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 43

d)

1766

0.2 KMU

2075 1992

1764

1556

1556

1800

1600

1558

1764

2000

1800

1600

Wavenumber / cm -1

11

ACS Paragon Plus Environment

32

Page 33 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

Figure 9. DRIFT spectra over different surface area Ru catalysts at different reaction times

2

(from bottom to top: 0, 1, 5, 9, 15, 75, 135,195, 255, 495, 615, and 735 min) in CO2-ref gas

3

mixture at 190°C. a) Ru/TiO2-1, b) Ru/TiO2-2, c) Ru/TiO2-3, and d) Ru/TiO2-4 (OCO stretch

4

vibration region).

5

The sets of spectra presented in Figure 9, which were recorded during 1000 min exposure to

6

CO2-ref reaction gas at 190 °C, show for all catalysts except Ru/TiO2-1 two dominant peaks in

7

the CO region, one at around 2075 cm-1 and one at around 2010 cm-1, where the former one

8

essentially disappears during time on stream. The band related to the Ru-multicarbonyl species

9

(~2075 cm-1) exhibits a small red shift, down to 2072 cm-1, compared to the SR-ref 6000 gas

10

mixture. This shift can be explained by the rather low coverage of COad resulting from the CO2

11

dissociation. The band at 2018 – 2009 cm-1 associated with on top adsorbed COad species shows

12

a similar red shift compared to the SR-ref 6000 gas mixture, it appeared in the range 2005 -1992

13

cm-1. Bands at 2009, 2058 and 1982 cm-1, which had been detected in the presence of CO in the

14

reaction gas mixture, did not appear in CO-free reformate gas.

15

Also in these measurements we will focus on the situation reached under steady-state conditions.

16

We will estimate the tendency for CO2 dissociation, which is directly correlated with the

17

selectivity for CO methanation, from the amount of COad present under steady-state conditions

18

during exposure to CO2-ref. This is determined from the intensity of the COad related band under

19

steady-state conditions, after normalization to the intensity obtained under saturation conditions,

20

during exposure to SR-ref 6000. The normalized intensities, which are presented in Figure 10,

21

clearly show that the formation of COad and Ru-carbonyl species from the dissociation of CO2

22

decreases with increasing support surface area and, hence, decreasing Ru particle size. The

23

relative intensities of the carbonyl related band decrease from 17% over Ru/TiO2-2 catalyst to

ACS Paragon Plus Environment

33

ACS Catalysis

1

0.4 % over Ru/TiO2-4. (Note that the normalization results in measurable contributions from the

2

Ru-multicarbonyl species, although on an absolute scale the band intensities are very low.)

3

Interestingly, the lowest surface area catalyst (Ru/TiO2-1) did not show any formation of Ru-

4

carbonyl species, which resembles the behavior in SR-ref 6000 (see section 3.3), where they

5

were also absent. For the on-top adsorbed COad species, the relative band intensity is about 125%

6

for the Ru/TiO2-1 catalyst, and it decreased with increasing support surface area, reaching 41%

7

for the Ru/TiO2-4 catalyst (see Figure10). Note that the relative intensity > 100% for COad

8

observed on the Ru/TiO2-1 catalyst is possible due to the much lower intensity for the bridge

9

bonded CO.

Normalized Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 43

a)

125 100

b)

75

c)

50

d)

25 x 10

0

0

50

100

150 2

SA / m g

10

x 10

200

250

-1

11

Figure 10. Normalized intensity of adsorbed surface species (empty columns: 2071 cm-1; filled

12

columns: 2009- 1999 cm-1) resulting from dissociation of CO2 in CO2-ref on a) Ru/TiO2-1, b)

13

Ru/TiO2-2, c) Ru/TiO2-3, and d) Ru/TiO2-4.

14

These trends in the coverages of COad and multicarbonyl species from CO2-ref agree well with

15

the observed increase of the selectivity with increasing surface area in SR-ref 100, which goes

16

along with a decreasing Ru particle size. Hence, also for the Ru/TiO2 catalysts the selectivity for

ACS Paragon Plus Environment

34

Page 35 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

CO methanation is correlated with their CO2 dissociation activity. Moreover, also in the present

2

case the activity for CO2 dissociation decreases with decreasing of Ru particle diameter, as

3

observed earlier for Ru/zeolite and Ru/Al2O3.11,12,17 Different from the latter cases, however, this

4

is not correlated with a stabilization of the dissociation product COad, as expected in a simple

5

picture based on the BEP principle. Hence, for catalysts with significant metal-support

6

interactions the physical origin for the increasing selectivity with decreasing Ru particle size

7

below ca. 2 nm may be different than in the previous cases. This is topic of ongoing studies in

8

our laboratory.

9

4 Conclusions

10

Employing a combination of kinetic and in situ spectroscopic measurements as well as detailed

11

ex situ catalyst characterization, we could show for the selective CO methanation reaction on a

12

set of Ru/TiO2 catalysts with rather similar Ru loading, Ru particle size, and TiO2 phase

13

composition that both the activity (Ru mass based activity and TOFs) and the selectivity of these

14

catalysts in two reaction mixtures with medium (0.6%) and very low (0.01 %) CO contents

15

depend sensitively on the BET surface area of the catalysts. The activity increases first up to a

16

maximum surface area of 121 m2 g and then decreases again for the catalyst with the highest

17

surface. The selectivity for CO methanation is 100% for reaction in reformate gases with

18

medium CO content, due to site blocking for CO2 dissociation by COad, it increases steadily for

19

reaction in reformate with very low CO content, where site blocking effects are little relevant,

20

reaching close to 100% for the catalyst with the highest surface area. The variation in reaction

21

behavior with increasing catalyst surface area goes along with and is affected by the following

22

effects:

ACS Paragon Plus Environment

35

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

i)

Page 36 of 43

An increasing dispersion / decreasing Ru particle size (with increasing surface area),

2

which similar to previous findings for other supported Ru catalysts (Ru/zeolite,

3

Ru/Al2O3) dominates the CO2 dissociation activity, results in an increasing selectivity

4

for CO methanation under conditions of low CO contents, where COad site blocking

5

effects are negligible (particle size effect).

6

ii)

The increasing Ru dispersion activity also causes the CO methanation activity to

7

increase, up to the Ru/TiO2-3 catalyst with a surface area of about 121 m2 gcat-1. With

8

further increasing surface area, counteracting strong metal-support interaction (SMSI)

9

effects overcompensate this by reducing the surface area available for reaction, as

10

evident from the COad uptake at saturation (steady-state in SR-ref 6000).

11

Accordingly, the activity drops sharply for the latter catalyst due to an SMSI effect.

12

iii)

Metal-support interactions are evidenced also by an increasingly stronger CO

13

adsorption on the Ru particles with increasing BET surface area. Considering that

14

with increasing surface area also the Ru particle size decreases, this variation of the

15

CO adsorption strength is in contrast to previous observations on other oxide

16

supported Ru catalysts (Ru/zeolite, Ru/Al2O3), where the CO adsorption energy was

17

found to decrease with increasing Ru dispersion / decreasing Ru particle size. This

18

observation points to an MSI effect for the CO adsorption energy which

19

overcompensates the particle size effects dominant in the latter catalysts.

20

iv)

Metal-support interactions must be responsible also for the disparate trends for CO

21

adsorption strength and for the selectivity for CO methanation / activity for CO2

22

dissociation, with an increasing selectivity for CO methanation / decreasing activity

ACS Paragon Plus Environment

36

Page 37 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

for CO2 dissociation together with an increasing CO adsorption strength with

2

increasing surface area / increasing Ru dispersion, where the latter is opposite to the

3

trend for the CO adsorption energy observed in previous studies on Ru/zeolite and

4

Ru/Al2O3 catalysts and opposite to the trend expected from simple arguments based

5

on the BEP principle.

6

Overall, the data clearly demonstrate the role of metal-support interactions in the selective CO

7

methanation on Ru/TiO2, which affect the reaction behavior in addition to the Ru particle size

8

effects well known for other supported Ru catalysts. Metal-support interactions also provide an

9

additional parameter to modify and steer the catalytic behavior of the Ru catalyst, which is

10

highly relevant also for further catalyst optimization for technical applications.

11

Supporting Information

12

Additional figures including kinetic, XRD, XPS and Infrared data in support of the manuscript.

13

This material is available free of charge via the Internet at http://pubs.acs.org.

14

AUTHOR INFORMATION

15

Corresponding Author

16

* Author to whom correspondence should be addressed, e-mail: juergen.behm@uni-ulm

17

Permanent Address

18

§

19

12613, Egypt

20

Author Contributions

Ali M. Abdel-Mageed: Department of Chemistry, Faculty of Science, Cairo University, Giza

ACS Paragon Plus Environment

37

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Page 38 of 43

The authors declare no competing financial interest.

2 3

ACKNOWLEDGMENT

4

Financial support by the ministry of research of Egypt and the German Academic Exchange

5

Service (DAAD) is gratefully acknowledged. Additionally, we thank Dr. Thomas Diemant, Ms.

6

C. Egger, and Ms. M. Lang (all Ulm University) for XPS measurements, BET measurements,

7

and ICP-OES measurements, respectively.

8

REFERENCES

9

(1)

10 11

2007, 326, 213. (2)

12 13

Dagle, A. R.; Wang, Y.; Xia, G.-G.; Strohm, J. J.; Holladay, J.; Palo, D. R. Appl. Catal. A

Chen, A.; Miyao, T.; Higashiyama, K.; Yamashita, H.; Watanabe, M. Angew. Chem. Int. Ed. 2010, 49, 9895.

(3)

14

Tada, S. and Kikuchi, R. Mechanistic study and catalyst development for selective carbon monoxide methanation. Catal.Sci.Technol. 2015.

15

(4)

Takenaka, S.; Shimizu, T.; Otsuka, K. Int. J. Hydrogen Energy 2004, 29, 1065.

16

(5)

Kowalczyk, Z.; Stolecki, K.; Rarog - Pilecka, W.; Miskiewicz, E.; Wiczkowska, E.;

17

Karpinski, Z. Appl. Catal. A 2008, 342, 35.

18

(6)

Eckle, S.; Anfang, H.-G.; Behm, R. J. Appl. Catal. A 2011, 391, 325.

19

(7)

Panagiotopoulou, P.; Kondarides, D. I.; Verykios, X. E. Appl. Catal. B 2009, 88, 470.

20

(8)

Mills, G. A.; Steffgen, F. W. Catal. Rev. Sci. Eng. 1974, 8, 159.

ACS Paragon Plus Environment

38

Page 39 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

(9)

Solymosi, F.; Erdóhelyi, A.; Kocsis, M. J. Catal. 1980, 65, 428.

2

(10)

Tóth, M.; Kiss, J.; Oszkó, A.; Pótári, G.; László, B.; Erdöhelyi, A. Top. Catal. 2012, 55,

3

747.

4

(11)

Eckle, S.; Augustin, M.; Anfang, H.-G.; Behm, R. J. Catal. Today 2012, 181, 40.

5

(12)

Abdel-Mageed, A. M.; Eckle, S.; Anfang, H.-G.; Behm, R. J. J. Catal. 2013, 298, 148.

6

(13)

Abdel-Mageed, A. M., Eckle, S., and Behm, R. J., J. Am. Chem. Soc., 2015, 137, 8672.

7

(14)

Panagiotopoulou, P.; Kondarides, D. I.; Verykios, X. E. Appl. Catal. A 2008, 344, 45.

8

(15)

Jiménez, V.; Sánchez, P.; Panagiotopoulou, P.; Valverde, J. L.; Romero, A. Appl. Catal.

9 10

A 2010, 390, 35. (16)

11 12

Tada, S.; Kikuchi, R.; Takagaki, A.; Sugawara, T.; Oyama, S. T.; Urasaki, K.; Satokawa, S. Appl. Catal. B 2013, 140-141, 258.

(17)

13

Abdel-Mageed, A. M., Widmann, D., Eckle, S., Behm, R. J., ChemSusChem 2015, in press, DOI : 10.1002/cssc.201500883.

14

(18)

King, D. L.; Peri, J. B. J. Catal. 1979, 79, 164.

15

(19)

Kellner, C. S.; Bell, A. T. J. Catal. 1982, 75, 251.

16

(20)

Che, M.; Bennet, C. O. Adv. Catal. 1989, 36, 55.

17

(21)

Rostrup-Nielsen, J. R.; Pedersen, K.; Sehested, J. Appl. Catal. A 2007, 330, 134.

18

(22)

Tada, S.; Kikuchi, R.; Urasaki, K.; Satokawa, S. Appl. Catal. A 2011, 404, 149.

ACS Paragon Plus Environment

39

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 40 of 43

1

(23)

Gupta, N. M.; Tripathi, A. K. Bull. Catal. Soc. India 2003, 213.

2

(24)

Solymosi, F.; Erdóhelyi, A.; Bansagi, T. J. Catal. 1981, 68, 371.

3

(25)

Solymosi, F.; Tombacz, I.; Kocsis, M. J. Catal. 1982, 75, 78.

4

(26)

Braunschweig, E. J.; Logan, A. D.; Datye, A. K.; Smith, D. J. J. Catal. 1989, 118, 227.

5

(27)

Bracey, J. D.; Burch, R. J. Catal. 1984, 86, 384.

6

(28)

Erdóhelyi, A.; Pasztor, M.; Solymosi, F. J. Catal. 1986, 98, 166.

7

(29)

Tauster, S. J.; Fung, S. C.; Garten, R. L. J. Am. Chem. Soc. 1978, 100, 170.

8

(30)

Tauster, S. J.; Fung, S. C.; Baker, R. T. K.; Horsley, J. A. Science 1981, 211, 1121.

9

(31)

Tauster, S. J. Acc. Chem. Res. 1987, 20, 389.

10

(32)

Uzio, D.; Berhault, G. Catal. Rev. 2010, 52, 106.

11

(33)

Hamadeh, I. M.; Griffiths, P. R. Appl. Spectrosc. 1987, 41, 682.

12

(34)

Miao, L.; Jin, P.; Kaneko, K.; Terai, A.; Nabatova-Gabain, N.; Tanemura, S. Appl. Surf.

13 14

Sci. 2003, 212-213, 255. (35)

15 16

Q. Nature 2008, 453, 638. (36)

17 18

Yang, H. G.; Sun, C. H.; Qiao, S. Z.; Zou, J.; Liu, G.; Smith, S. C.; Cheng, H. M.; Lu, G.

Elmasides, C.; Kondarides, D. I.; Grünert, W.; Verykios, X. E. J. Phys. Chem. B 1999, 103, 5227.

(37)

Castillo, R.; Koch, B.; Ruiz, P.; Delmon, B. J. Catal. 1996, 161, 524.

ACS Paragon Plus Environment

40

Page 41 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

(38)

Kruse, N.; Chenakin, S. Appl. Catal. A 2011, 391, 367.

2

(39)

Eckle, S.; Denkwitz, Y.; Behm, R. J. J. Catal. 2010, 269, 255.

3

(40)

Jacobsen, C. J. H.; Dahl, S.; Hansen, P. L.; Törnqvist, E.; Jensen, L.; Topsøe, H.; Prip, D.

4 5

V.; Møenshaug, P. B.; Chorkendorff, I. J. Mol. Catal. A 2000, 163, 19. (41)

6 7

Honkala, K.; Hellman, A.; Remediakis, I. N.; Logadottir, A.; Carlsson, A.; Dahl, S.; Christensen, C. H.; Nørskov, J. K. Science 2005, 307, 555.

(42)

Andersson, M. P.; Abild-Pedersen, F.; Remediakis, I. N.; Bligaard, T.; Jones, G.;

8

Engbæk, J.; Lytken, O.; Horch, S.; Nielsen, J. H.; Sehested, J.; Rostrup-Nielsen, J. R.;

9

Nørskov, J. K.; Chorkendorff, I. J. Catal. 2008, 255, 6.

10

(43)

King, D. L. J. Catal. 1978, 51, 386.

11

(44)

Abdel-Mageed, A. M., Eckle, S., D. Widmann, Behm, R. J., submitted for publication.

12

(45)

Evans, M. G.; Polanyi, N. P. Trans. Faraday Soc. 1938, 34, 11.

13

(46)

Miyao, T.; Shen, W.; Chen, A.; Higashiyama, K.; Watanabe, M. Appl. Catal. A 2014,

14

468, 187-192.

15

(47)

Djinovic, P.; Galletti, C.; Specchia, S.; Specchia, V. Catal. Today 2011, 164, 282-287.

16

(48)

Djinovic, P.; Galletti, C.; Specchia, S.; Specchia, V. Top. Catal. 2011, 54, 1042-1053.

17

(49)

Eckle, S.; Anfang, H.-G.; Behm, R. J. J. Phys. Chem. C 2011, 115, 1361.

18

(50)

Chin, S. Y.; Williams, C. T.; Amiridis, M. D. J. Phys. Chem. B 2006, 110, 871.

ACS Paragon Plus Environment

41

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 42 of 43

1

(51)

Guglielminotti, E.; Zecchina, A.; Bossi, A.; Camia, M. J. Catal. 1982, 74, 240.

2

(52)

Gupta, N. M.; Londhe, V. P.; Kamble, V. S. J. Catal. 1997, 169, 423.

3

(53)

Solymosi, F.; Raskó, J. J. Catal. 1989, 15, 107.

4

(54)

Panagiotopoulou, P.; Kondarides, D. I.; Verykios, X. E. J. Phys. Chem. C 2011, 115,

5

1220.

6

(55)

Londhe, V. P.; Kamble, V. S.; Gupta, N. M. J. Mol. Catal. A 1997, 121, 33.

7

(56)

Guglielminotti, E.; Bond, G. C. J. Chem. Soc. Faraday Trans. 1 1990, 86, 979.

8

(57)

Robbins, J. L. J. Catal. 1989, 115, 120.

9

(58)

Dalla Betta, R. A. J. Phys. Chem. 1975, 79, 2519.

10

(59)

Martins, R. L.; Baldanza, M. A. S.; Schmal, M. J. Phys. Chem. B 2001, 105, 10303.

11

(60)

Mizushima, T.; Tohji, K.; Udagawa, Y.; Ueno, A. J. Phys. Chem. 1990, 94, 4980.

12

(61)

Gupta, N. M.; Kamble, V. S.; Kartha, V. B.; Iyer, R. M.; Thampi, K. R.; Gratzel, M. J.

13

Catal. 1994, 146, 173.

14

(62)

Inderwildi, O. R.; Jenkins, S. J.; King, D. A. J. Phys. Chem. C 2008, 112, 1305.

15

(63)

Fajín, J. L. C., Gomes, J. R. B., Cordeiro, M. N. D. S. J. Phys. Chem. C 2015, 119,

16 17 18

16537. (64)

Prairie, M. R.; Renken, A.; Highfield, J. G.; Thampi, K. R.; Grätzel, M. J. Catal. 1991, 129, 130.

ACS Paragon Plus Environment

42

Page 43 of 43

1

(65)

Baddour, R. F.; Deibert, M. C. J. Phys. Chem. 1966, 70, 2173.

2 3 4 5 6 7 8 9 10

Table of Contents:

CO + 3 H2 C

Ru/TiO2

CH4 + H2O

x

O

Activity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

x

Ru

x

TiO2

C Ru

x 11

O

TiO2

TiO2 surface area

12

ACS Paragon Plus Environment

43