CO2 reduction on supported Ru/Al2O3 catalysts: Cluster size dependence of product selectivity

Article abstract of DOI:10.1021/cs400381f

The catalytic performance of a series of Ru/Al₂O₃ catalysts with Ru content in the 0.1-5% range was examined in the reduction of CO₂ with H₂. At low Ru loadings (≤0.5%) where the active metal phase is highly dis

Full text of DOI:10.1021/cs400381f

Subscriber access provided by UNIVERSITY OF CALGARY
Article
CO2 reduction on supported Ru/Al2O3 catalysts:
cluster size dependence of product selectivity
Ja Hun Kwak, Libor Kovarik, and Janos Szanyi
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/cs400381f • Publication Date (Web): 16 Sep 2013
Downloaded from http://pubs.acs.org on September 27, 2013
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 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
CO reduction on supported Ru/Al O catalysts: cluster size dependence of
2
2
3
product selectivity
1
#
2
1*
Ja Hun Kwak * , Libor Kovarik , and János Szanyi
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
2
Institute for Integrated Catalysis, Environmental Molecular Sciences Laboratory, Pacific
Northwest National Laboratory, Richland, WA 99352, USA
Corresponding authors: jhkwak@unist.ac.kr; janos.szanyi@pnnl.gov
#
Current Address: School of Nano-Bioscience and Chemical Engineering, UNIST
1
00 Banyeon-Ri, Ulsan 689-798, Korea:
Abstract
The catalytic performance of a series of Ru/Al O catalysts with Ru content in the 0.1ꢀ5% range
2
3
was examined in the reduction of CO with H . At low Ru loadings (≤0.5 %) where the active
2
2
metal phase is highly dispersed (mostly atomically) on the alumina support CO is formed with
high selectivity. With increasing metal loading the selectivity toward CH formation increases,
4
while that for CO production decreases. In the 0.1% Ru/Al O catalyst Ru is mostly present in
2
3
atomic dispersion as STEM images obtained from the fresh sample prior to catalytic testing
reveal. STEM images recorded form this same sample following temperature programmed
reaction test clearly show the agglomeration of small metal particles (and atoms) into 3D
clusters. The clustering of the highly dispersed metal phase is responsible for the observed
dramatic selectivity change during elevated temperature tests: dramatic decrease in CO, and
large increase in CH selectivity. Apparent activation energies, estimated from the slopes of
4
Arrhenius plots, of 82 kJ/mol and 62 kJ/mol for CO and CH formation were determined,
4
respectively, regardless of Ru loading. These results suggest that the formation of CO and CH4
follow different reaction pathways, or proceed on active centers of different nature. Reactions
with CO /H and CO/H mixtures (under otherwise identical reaction conditions) reveal that the
2
2
2
onset temperature of CO reduction is about 150 ºC lower than of CO reduction.
2
Keywords: CO reduction; Ru/Al O ; CO/CH selectivity; Ru dispersion; reaction mechanism
2
2
3
4
1
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Introduction
The conversion of CO to high energy density organic molecules (e.g., methanol or
2
methane) has been proposed over both homogeneous and heterogeneous catalysts containing
transitions metals (e.g., Cu, Ni and Pd) as active centers [1,2,3]. Considerable work has been
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
aimed at designing and synthesizing heterogeneous CO reduction catalysts [4,5,6,7,8]. The
2
activity and selectivity of these catalysts have been shown to be very sensitive to the cluster size
and the shape of the metal particles dispersed on the support, as well as to the interaction
between the active metals and oxide supports [9, 10]. However, these features of oxideꢀ
supported metal catalysts remain, to a great extent, hard to control due to the nature of synthetic
protocols, and, in particular, to sintering of metal clusters during the activation process. The
heterogeneous catalytic conversion of CO is currently not feasible due to the demanding
2
reaction conditions (e.g., high catalyst bed temperature) originating from the chemical inertness
of CO . Therefore, understanding the elementary reaction steps of catalytic CO reduction is
2
2
critical in order to design economically viable catalytic systems.
Despite the ongoing research efforts, the role of support and the control of CO/CH4
selectivity in the reduction of CO with H have not been well established, in particular not on
2
2
subꢀnanometerꢀsized supported metal atoms/clusters. Recently we have reported on the very
unique catalytic properties of isolated Pd atoms in the reduction of CO [11]. In that report, we
2
showed that atomically dispersed supported metals can be catalytically active even in the
demanding reaction of CO reduction but their activity and selectivity patterns differ by a large
2
extent from those of 3D metal particles. The results of CO hydrogenation reaction on Pd/Al O
3
2
2
and Pd/MWCNT catalysts have unambiguously proved the need of two different functionalities
in an active catalyst. The reduction of CO requires the presence of a catalyst component that is
2
able to activate CO (either the support oxide (Al O ) or an oxide promoter), and a metallic
2
2
3
component (here Pd) that is able to dissociate H . When both of these functionalities are present,
2
the CO and CH selectivities seem to be determined by the sizes of the metal particles.
4
In this study we prepared a series of Ru on alumina catalysts with Ru loadings that
assured a dispersion range from atomic to 3D clusters and tested their CO reduction
2
performances. Atomically dispersed Ru on alumina initially produced CO exclusively by CO2
reduction. This is in complete contrast with the catalytic behavior of 3D Ru clusters supported on
alumina that have been known to be efficient methanation catalysts [12,13,14]. Based on these
2
ACS Paragon Plus Environment

Page 3 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
results we propose that CO can be produced by different reaction mechanisms on active Ru
centers with different particle sizes. Furthermore, CO may not be a simple reaction intermediate
in the path of CO hydrogenation to CH .
2
4
Experimental:
Al O –supported, 0.1%, 0.5%, 1%, 2% and 5 wt% Ru catalysts were prepared on a
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
2
3
2
commercial γꢀAl O powder (Condea, BET surface area = 200 m /g) by the incipient wetness
2
3
method using ruthenium(III) nitrosyl nitrate solution (Ru(NO)(NO ) (OH) , x+y =3 in dilute
3 x
y
nitric acid, 1.5% Ru, Aldrich) as the precursor.
CO reduction activity measurements were conducted by a temperature programmed reaction
2
methods in a packed bed reactor using 50 mg of catalyst powder samples (quartz reactor O.D. =
o
1
/2"). The catalysts were activated prior to catalytic measurements by calcination at 500 C for 2
o
h under 6.7 % O /He (flow rate = 60 ml/min) and followed by reduction at 500 C for 30 min
2
under 15 % H /He (flow rate = 60 ml/min). The activity was measured using a feed gas mixture
2
containing 5 % CO and 15% H in He (total flow rate = 60 ml/min and H /CO = 3; mcatalyst= 50
2
2
2
2
mg)). The concentrations of all reactant and product species were measured by a gas
chromatograph (HP 7820), with separation using capillary column (Supelco, Carboxeneꢀ1006
PLOT, 30m x 0.53mm I.D.) and a thermal conductivity detector. Temperature programmed CO
hydrogenation reaction was also performed using 5% CO, with the same protocols.
o
Catalytic activity changes by sintering were tested by timeꢀonꢀstream measurements at 350 C
up to 16 hrs on the 0.1% Ru/Al O catalyst prepared by the same pretreatments under the same
2
3
reaction conditions as described previously. Steady state activities were measured at 280 ~
o
3
20 C, which shown stable activities up to 1 h, with different Ru loaded catalysts using newly
activated samples for each temperature and initial activities were extrapolated. Catalytic
activities were evaluated under conditions where CO conversion stayed below 5% . Turnꢀoverꢀ
2
frequencies (TOF, number of CO converted/Rusurfaceꢁs) were calculated based on the number of
2
surface Ru atoms (as determined from H chemisorption measurements) on alumina. For the H2
2
chemisorption experiments the catalysts were pretreated under identical conditions that were
applied prior to the catalytic tests. The amount of chemisorbed H was determined on the thus
2
treated samples using a Micromeritics AutoChem 2920 Chemisorption Analyzer apparatus. The
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
o
activated catalyst (calcined then reduced) was held a 100 C during H chemisorption
2
measurements.
Highꢀresolution TEM imaging was performed with a FEI Titan 80ꢀ300 microscope operated
at 300 kV. The instrument is equipped with a CEOS GmbH doubleꢀhexapole aberration
corrector for the probeꢀforming lens, which allows imaging with 0.1 nm resolution in scanning
transmission electron microscopy (STEM) mode. The images were acquired in high angle
annular dark field (HAADF) with an inner collection angle of 52 mrad. The sample preparation
for the TEM measurements involved mounting of the powder samples on lacey carbon TEM
grids, and immediate loading into the TEM airlock to minimize extended exposure to
atmospheric O2.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Results and Discussion
CO conversion (panel a) and yields of CO (panel b) and CH (panel c) obtained in the
2
4
temperature programmed CO reduction experiments on 0.1, 0.5, 1, 2 and 5 wt% Ru/Al O
3
2
2
samples are displayed in Figure 1. On the 5% Ru/Al O catalyst, the CO reduction starts at ~
2
3
2
o
o
1
50 C and shows the maximum conversion at ~ 320 C and then the conversion decreases with
o
further temperature increase to ~500 C. Above 500 ºC the CO conversion increases again.
2
With decreasing Ru loading the CO conversion profiles are shifted toward higher temperature.
2
On the 0.1 wt% Ru/Al O sample, CO reduction starts above ~ 300ºC and monotonically
2
3
2
increases, then slows down above 450 ºC. CO conversion levels above this temperature
2
(~450ºC) are practically the same for all Ru loadings studied, although the loading changes by a
factor of 50. Even though CO formation starts at practically the same temperature (~ 300ºC) for
all Ruꢀloaded samples, the activity profiles were very different. Interestingly, catalysts with
lower amount of Ru showed significantly higher rate of CO formation at relatively low
temperature, while over Ru/Al O catalysts with Ru loading ≥2 % the CO yield increased
2
3
monotonically. The highest CO yield in the 350ꢀ500 ºC temperature range was observed for the
lowest Ruꢀloaded sample (0.1%). The CO yield increased rapidly after the onset temperature of
~
350 ºC, but then it leveled off between 450 and 500 ºC, and at even higher temperatures it
followed exactly the same path we observed for all the other catalysts. When the Ru loading
increased to 0.5% the trend in the CO yield was similar to that observed for the 0.1% sample,
however, its initial high activity leveled off at a much lower CO yield than that seen for the 0.1%
catalyst. For the 1% sample the high CO yield region at low temperature is even less evident
4
ACS Paragon Plus Environment

Page 5 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
than over the other two low Ruꢀloaded samples, but it is still clearly distinguishable. Above 500
o
C the CO yield profiles are identical for all catalysts studied, and practically fall on a single line.
The CH yield profiles indicate that methane formation rate increases with increasing Ru loading
4
and, at the same time, the temperature where the maximum methane yield is observed shifts to
lower values with increasing Ru loading. These temperature programmed reaction results were
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
completely consistent with our previously reported results on CO reduction on Pd/alumina,
2
where we have unambiguously showed that isolated, single Pd atoms favor the formation of CO,
while methane formation is prevalent on Pd clusters. These results are very peculiar in light that
Ru has been well known as a very efficient methanation catalyst [15]. The CO conversion as
2
well as the CO and CH yields trends seem to suggest the existence of three different reaction
4
regimes as the temperature is increased from 25 to 500 ºC. The formation of CH is prevalent at
4
low reaction temperatures, and the CH yield increases proportionally with Ru loading. This is
4
in line with the conclusions of earlier studies that CH formation is favored on large metal
4
clusters. With increasing Ru loading the size of metal clusters increases, as it has been
substantiated by TEM measurements. At very low Ru loading (0.1 wt%) (highly dispersed Ru)
no CH formation is seen even at temperatures (~350 ºC) where the CH yield reached its
4
4
maximum at high Ru loadings. As we will show in the following paragraph, the low Ruꢀloaded
catalyst exhibits CH formation activity only after the onset of metal sintering, i.e., after the
4
formation of 3D metal clusters. The maximum in the CO conversion vs. temperature plot can
2
solely be attributed to the maximum in CH formation rate. This behavior suggests that at high
4
temperature the concentration of one of the reactants (either CO or H ) on the catalyst surface
2
2
decreases which leads to a drop in CO conversion, and also in the CH yield. In the second
2
4
temperature regime the dominant reaction is the formation of CO on highly dispersed Ru
particles. At low Ru loadings we see a particular shape of the CO yield traces as a function of
temperature. The onset temperature of CO formation at low Ru loadings is around 320 ºC. At
the lowest Ru loading (0.1 wt%) the CO yield increases fast with temperature, levels off around
450 ºC and then increases again above 500 ºC. With increasing Ru laoding, however, the CO
yield drops, and over the 5 wt% Ruꢀloaded sample we observe no CO formation. CO is only
produced on this sample as a result of a secondary reaction at this high temperature between CH4
formed on Ru clusters and the reactant CO (dry reforming of CO ). (The observed increase in
2
2
CO yield above 500 ºC over the low Ru loaded samples can also be attributed to the onset of this
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
secondary reaction.) The increase in CO conversion at high temperatures (above 500 ºC) on all
2
samples studied here can be attributed to the reaction between CH and CO .
4
2
CO methanation over supported Ru catalysts has been shown to be structure sensitive as it
x
depended on the particle size of Ru clusters [16-21]. For CO hydrogenation Che et al. [21]
observed that the TOF decreased with increasing Ru dispersion, and similar findings for CO2
methanation was reported by Kowalczyk et al [16] (although, in the latter study the size of Ru
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
clusters was shown to affect the CO hydrogenation rate in a lesser extent that the CO
2
hydrogenation). The variation in the methanation rate with Ru particle size might be (at least in
part) correlated with the effect of the support on the metal particles. The electronic state (charge
density) of the Ru particles on a given support material (here Al O ) my vary with particle size,
2
3
as it was proposed by Scire at al. [22], who reported that the electronic state of the active metal
(Ru) strongly influenced the CO selectivity in the CO hydrogenation process. They proposed
2
that a more oxidized Ru surface led to higher methanation activity, due to the higher hydrogen
and lower CO coverage. In order to vary the electronic properties of Ru particles supported on
YSZ solid electrolyte pellets Theleritis et al. [23] used the electrochemical promotion of catalysis
(EPOC) effect by applying external potential to the catalytic system. They found that at low
temperatures (up to 240 ºC) CO hydrogenation reaction yielded CH primarily, while CO was
2
4
2
ꢀ
the dominant product at high temperatures. Electrochemical O supply to the Ru particles
resulted in large increases in both the formation rate and selectivity of CH , and concomitantly a
4
large decrease in the CO formation rate.
The results of these temperature programmed reaction measurements may also be argued on
the basis of having more Ru in the catalyst translates to higher hydrogenation activity and,
therefore, extensive CH formation. This interpretation, however, cannot explain the formation
4
of larger amount of CO on catalysts containing less amount of Ru. In order to address this issue,
we performed isothermal CO hydrogenation on the 0.1% Ru/Al O catalyst at 350 ºC. At this
2
2
3
temperature single Ru atoms present on the alumina support initially show sintering during the
course of reaction. The initial selectivity of CO formation at this catalyst bed temperature is
much higher (over 84%) than that of CH production on the 0.1% Ru/Al O catalyst, as shown in
4
2
3
Figure 2(a). With increasing timeꢀonꢀstream, CO production rate increased slowly, while the
CH formation rate increased much faster and after ~ 200 min timeꢀonꢀstream, CO and CH were
4
4
produced in almost the same amounts. After 16 hrs timeꢀonꢀstream, CH production far
4
6
ACS Paragon Plus Environment

Page 7 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
exceeded the formation of CO, as the CO selectivity dropped to ~36%. These results indicate
that the Ru/Al O catalysts with low Ru loadings are not stable under the reducing reaction
2
3
conditions of CO reduction. In order to compare the activities of fresh and used catalysts, we
2
evaluated the steady state activities of two 0.1% Ru/Al O catalysts at 300 ºC: a freshly prepared
2
3
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
catalyst and the one that was reactivity tested at 350 ºC for 16hrs. The fresh catalyst showed
stable steady state activity at this lower reaction temperature of 300 ºC in comparison to that
observed at 350 ºC. As the results of Figure 2(b) substantiate, the overall CO conversion
2
activity of the 350 ºCꢀtested catalyst increased ~ 4.5 times and most of the activity increase came
from the ~ 20 fold increase in the CH production rate. At the same time, the increase in CO
4
TOF at steady state at 300 ºC over the 350 ºCꢀtested sample was much smaller, only about 1.5X.
A likely explanation for the observed variation in catalytic activity at high temperature with
timeꢀonꢀstream is the change in metal particle size under reaction conditions. As we have seen
for the Pd/Al O catalysts [11] both the overall catalytic activity (i.e., CO conversion) and the
2
3
2
product (CO and CH ) selectivity were strongly dependent on the active metal particle size. In
4
order to visualize the change in metal particle size (changes in Ru dispersion), we collected
STEM images from two catalyst samples: form a freshly prepared 0.1% Ru/Al O catalyst and
2
3
from the one that was reactionꢀtested at 350 ºC. The freshly prepared 0.1% Ru/Al O sample
2
3
showed almost exclusively atomically dispersed Ru (and very small Ru aggregates) on the
alumina support (image a in Fig. 3). On the contrary, after reaction test at 350 ºC for 16 hrs,
mostly nmꢀsized (some of them even larger than 5 nm) 3D Ru clusters are visible in the STEM
images (image c in Fig.3). These results strongly support our hypothesis that CO formation is
favored on single Ru atoms supported on alumina, while Ru clusters are mostly active for CH4
formations. Therefore, we conclude that the high methanation activity reported for Ru on
alumina catalysts originate from large (3D) Ru clusters. The catalytic data discussed in the
previous section seem to suggest that the reaction mechanism is different for the small
(atomically dispersed) and larger (3D) Ru particles. Therefore, next we investigated how the
reaction mechanisms changed with the variation of active metal particle size. To this end, we
performed kinetic measurements on all the Ruꢀloaded alumina supported catalysts prepared
(
from 0.1 to 5% Ru loading). A series of Arrhenius plots of CO TOF (calculated based on the
2
number of surface Ru atoms on the alumina supported metal clusters) are displayed in Fig. 4(a),
each showing reasonably linearity in temperature regime explored in this study (270ꢀ350 ºC).
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The CO TOF is the lowest over the 0.1wt% Ru/alumina catalyst, and increases dramatically
2
over the 0.5 wt% sample, and further over the 1 wt% catalyst. Above 1 wt% Ru loading the CO2
TOF increases only marginally.. Good linearity in the Arrhenius plots for both CO and methane
TOFs is shown in Figure 4(b) and (c), respectively, except for the 0.1% Ru/Al O sample which
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
is related to the significant error in the calculation of TOF due to low CH selectivity. At Ru
4
loadings up to 2 % the CO TOF values are very similar, and only drop significantly on the
highest Ru loaded sample (5 wt%), as large metal clusters tend to produce no CO. The TOF
plots of CH at Ru loadings higher than 1% are very close to each other, but there is a systematic
4
increase in CH TOF with increasing Ru loading. This observation suggests that Ru clusters
4
exhibit very similar catalytic activities toward CH formation above a certain cluster size. These
4
results are consistent with our previous interpretations that single Ru atoms or interfacial Ru
favor CO formation, while Ru clusters favor CH formation. From the slopes we can estimate
4
the apparent activation energies of CO and CH formation. Interestingly, the activation energy
4
for CO formation (~82 kJ/mol) was always ~20 kJ/mole higher than that for CH formation (~62
4
kJ/mol), regardless of Ru loading. If one considers CO as an intermediate in the path toward
CH formation in the CO hydrogenation process, the activation energy for CO formation should
4
2
be equal to or lower than that for CH formation. Therefore, we propose that CO is either
4
formed by a different route (not as an intermediate in the reduction toward CH ) or on a different
4
active center than CH . The E for methane formation obtained from the Arrhenius plots in this
4
a
study is comparable to those reported previously by Solymosi et at. [24,25] for CO methanation
2
over 5% Ru/Al O (67.4 kJ/mol), and by Bartholomew et al. [26] for Ru/SiO (72 kJ/mol).
2
3
2
Changes in the apparent activation energy with reaction conditions applied has also been
reported in the latter publication, due, possibly, to the variation in the rate determining step under
different reaction conditions.
In Figure 5(a), we show CO and CH TOFs measured at 300 ºC as a function of Ru loading in
4
the reduction of CO . With increasing Ru loading the trends in CO and CH TOFs are opposite:
2
4
the TOF for CO formation is practically constant up to 2 wt% Ru loading and then it decreases,
while that for CH formation increased in the entire Ru loading regime studied (0.1 – 5 wt%).
4
The CO selectivity below 1% Ru on alumina is evidently increases sharply with the decrease in
Ru loading (Fig. 5b). We believe that these results indicate high Ru dispersion at loadings less
than 1%, when most of the Ru is present as isolated single atoms and small metal aggregates, and
8
ACS Paragon Plus Environment

Page 9 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
the production of CO is favored. On the other hand, at Ru loadings high enough to form Ru
clusters (starting from ~1%), the reduction of CO proceeds all the way to CH . These results
2
4
are fully consistent with our previous report on the reactivity of isolated Pd atoms that have been
shown to favor the formation of CO, while mostly CH formed on the Pd clusters [11]. The
4
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
results of both of these studies strongly suggest that the active sites for CO formation are highly
dispersed single metal atoms (these form CO with high selectivity), and possibly small metal
aggregates, as well. However, the formation of CH can only proceed in the presence of metal
4
clusters that are able to supply large amounts of atomic hydrogen to this process. As we have
mentioned above, the apparent activation energy for CH formation is about 20 kJ/mol lower
4
than that of CO formation over every Ru/Al O catalyst studied here. This result seems to
2
3
suggest that CH production may not proceed through an intermediate that formed from CO
4
produced in the initial reduction step. To test this we performed temperature programmed
reduction of both CO and CO on freshly prepared 0.1% Ru/Al O catalysts. As the results
2
2
3
displayed in Figure 6 show, the onset temperature for CO hydrogenation on the 0.1% Ru/Al O
3
2
2
is ~300ꢀ350 ºC, while that for CO hydrogenation is ~500 ºC. We propose that this ~150 ºC
upward shift in the onꢀset temperature for hydrogenation provides two important insights into the
reaction mechanisms. First, CO hydrogenation to CO might proceed by two different, in
2
parallel occurring mechanisms: hydrogenation on metallic surfaces and reverse water gas shift
reaction. Second, alumina supported isolated single Ru atoms can catalyze CO reduction to CO
2
but cannot catalyze further hydrogenation to CH due, perhaps, to their limited hydrogen
4
activation functionality. This would be consistent with our previous finding [11] in that dual
catalyst functionality was required for CO hydrogenation to CH over supported Pd catalysts.
2
4
In a Ruꢀdoped ceria catalyst Metiu et al., [27] have also found that “gasꢀphase CO is not a
reaction intermediate for the methanation of CO by Ce0.95Ru0.05O ”. In a recent study Ussa
2
2
Aldana et al. carried out in operando FTIR measurements on the CO methanation reaction over
2
Ni/CeZrO catalysts [28]. Their results unambiguously show that methane formation does not go
2
through a CO intermediate, rather surface carbonate/formate species formed on the oxide support
play a critical role in the formation of CH . Adsorbed CO is proposed to hydrogenate stepwise,
4
2
forming bicarbonates, formates, formaldehyde, and finally methoxide. Supported Ni clusters
dissociate H and supply H atoms for these hydrogenation steps. The mechanism of CO2
2
methanation on supported Ru catalysts has also been strongly debated, and no clear consensus on
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the actual reaction path has been reached. Based on the results of a DRIFT spectroscopy study
Prairie et al. proposed formic acid as a critical intermediate in the CO methanation process [29].
2
Subsequently, the role of formic acid or formate intermediate was questioned by Traa et al., who
found in their kinetic study on CO methanation ovet Ru/TiO catalysts that CH formation
2
2
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
proceeded through the hydrogenation of surface carbon as the rate determining step [30] Our
current kinetic and spectroscopy studies are focused on understanding the reaction mechanisms
for the formation of both CO and CH over these Ruꢀbased catalysts at different Ru loadings, and
4
identifying active catalytic centers of these materials and key reaction intermediates.
Conclusion:
The reactivities of Ru/Al O catalysts in the Ru loading range of 0.1ꢀ5wt% were studies in the
2
3
reduction of CO with H . The main focus of the work was to find correlation between the active
2
2
metal dispersion and the catalytic performance of these materials. At very high metal dispersion
metals mostly present in atomic dispersion, as evidence by STEM) the catalyst produces CO
(
with high selectivity. As 3D metal clusters form at higher Ru loadings (at and above 1 wt%) or
as a results of sintering, the selectivity toward CH formation increases significantly. Catalysts
4
with low metal loading, however, are unstable under reaction conditions of CO reduction, and
2
form large metal clusters. This clustering is accompanied in a large increase in CH selectivity
4
and drop in CO formation selectivity. Apparent activation energies of 82 and 62 kJ/mol were
estimated from the slopes of Arrhenius plots for CO and CH formation, respectively. The
4
difference in activation energy of ~20 kJ/mol for each catalyst studied suggests either different
reaction paths for the formation of CO and CH , or/(and) different active sites. The higher
4
apparent activation energy for CO formation also seems to suggest that CO is not an intermediate
in the formation of CH . This is further supported by the results of activity measurements over a
4
0.1% Ru/Al O catalyst, that showed about 125 ºC onset temperature difference between the CO
2
2
3
(~325 ºC) and CO (~450 ºC) reduction.
Acknowledgements
We thank Dr. Feng Gao for carrying out the H chemisorption measurements on all the
2
Ru/Al O catalysts discussed in this work. The catalyst preparation and catalytic measurements
2
3
were supported by a Laboratory Directed Research and Development (LDRD) project, while the
TEM work was supported by the Chemical Imaging Initiative at the Pacific Northwest National
1
0
ACS Paragon Plus Environment

Page 11 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
Laboratory (PNNL). PNNL is operated for the US Department of Energy by Battelle Memorial
Institute under contract number DEꢀAC05ꢀ76RL01830. JHK also acknowledges the support of
this work by the 2013 Research Fund of UNIST (Ulsan National Institute of Science and
Technology, Ulsan, Korea).
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
References:
1. Boffa, A.; Lin, C.; Bell, A.T.; Somorjai,. G.A., J.Catal., 1994, 149, 149.
2. Wambach, J.; Baiker, A.; Wokaun, A. Phys.Chem.Chem.Phys., 1999, 1, 5071.
3. R.P.A. Sneeden, J.Mol.Catal., 1982, 17, 349.
4. Fisher, I.A.; Bell, A. T. J.Catal., 1996, 162, 54.
5. Baiker, A. Appl.Organomet.Chem., 2000,14, 751.
6. Peterson, A. A.; AbildꢀPedersen, F.; Studt, F.; Rossmeisl, J.; Norskov, J.K. Energy &
Environ.Sci., 2010, 3, 1311.
7
. Durand, W.J.; Peterson, A.A.; Studt, F.; AbildꢀPedersen, F.; Norskov, J.K. Surf.Sci., 2011,
05, 1354.
8. Ferrin, P.; Mavrikakis, M. J.Am.Chem.Soc., 2009, 131, 14381.
6
9
1
1
1
. Clarke, D. B.; Bell, A.T.; J.Catal., 1995, 154, 314.
0. Clarke, D.B.; Suzuki, I.; Bell, A.T.; J.Catal., 1993, 142, 27.
1. Kwak, J.H.; Kovarik, L.; Szanyi, J. ACS Catalysis, in press,DOI: 10.1021/cs4001392.
2. Solymosi, F.; Erdıhelyi, A.; Kocsis, M.; J.Chem.Soc.-Faraday Trans., 1981, 77 1003.
13. Elliott, D.J.; Lunsford, J.H.; J.Catal., 1979, 57 11.
1
1
1
4. Kusmierz, M.; Catal. Today, 2008, 137 429.
5. Mills, G.A.; Steffgen, F.W. Catal. Rev., 1973,8 159.
6. Kowalczyk Z., Stolecki, K., RarogꢀPilecka, W., Moskiewicz, E., Wilczkowska, E.,
Karpinsky, Z., Appl.Catal A: General, 2008, 342, 35.
1
1
1
2
7. King, D.L., J.Catal., 1978, 51, 386.
8. Kellner, S.C., Bell, A.T., J.Catal., 1982, 75, 251.
9. Guczi L., Schay, Z., Matusek, K., Bogyai, I., Appl.Catal., 1986, 22, 289.
0. Venter, J.J., Vannice, M.A., Inorg.Chem., 1989, 28, 1634.
21. Che M., Bennett C.O., Adv.Catal., 1989, 36, 55.
1
1
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2. Scire, S., Crisafulli, C., Maggiore, R., Minico, S., Galvagno, S., Catal.Lett., 1998, 51,
4
1.
23. Theleritis, D., Souentie, S., Siokou, A., Katsaounis, A., Vayenas, C.G., ACS Catal., 2012,
, 770.
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
4. Solymosi, F. and Erdıhelyi, A. J. Mol. Catal., 1980, 8,471.
5. Solymosi F., Erdıhelyi, A.,Bánsági, T. J. Catal., 1981,68, 371.
2
26. Weatherbee G.D.; Bartholomew, C.H. J.Catal. 1984, 87, 352ꢀ362.
27. Sharma, S.; Hu, Z.; Zhang, P.; McFarland, E.W.; H. Metiu J.Catal. 2011, 278 297.
2
2
3
8. Ussa Aldana, P.A.; Ocampo, F.; Kolb, K.; Louis B.; ThibaultꢀStarzyk, F.; Daturi, M.;
Bazin, P.; Thomas, S.; Roger, A.C. Catal.Today, DOI: 10.1016/j.cattod.2013.02.019.
9. Prairie, M.R., Renken, A., Highfield, J.G., Thampi, K.R., Grätzel, M., J.Catal., 1991,
129, 130.
0. Traa, Y., Weitkamp, J., Chem.Eng.Technol., 1999, 21, 4.
1
2
ACS Paragon Plus Environment

Page 13 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
Figure captions
Figure 1. Temperature programmed CO reduction reaction on Ru/Al O catalysts: CO
2
2
2
3
conversion (a), CO (b) and CH (c) yields . (mcatalyst= 50 mg, 5% CO +15% H in He
4
2
2
(total flow rate = 60 ml/min), heating rate = 5 ºC /min)
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
Figure 2. (a) TOFs of CO , CO and CH as a function of timeꢀonꢀstream at 350 ºC over a 0.1%
2
4
Ru/Al O catalyst. (b) steadyꢀstate TOFs for CO conversion and CO/CH production over a
2
3
2
4
fresh and reꢀactivated (after reaction at 350 ºC) 0.1% Ru/Al O catalyst at 300 ºC.
2
3
Figure 3. STEM images from a 0.1% Ru/Al O catalyst before (images a and b) and after
2
3
(
images c and d) CO reduction at 350 ºC for 16 hrs.
2
Figure 4. Arrhenius plots for CO conversion (a), as well as for CO (b) and CH (c) formations
2
4
over Ru/Al O catalysts between 270 ºC and 350 ºC.
2
3
Figure 5. (a) TOFs of CO and CH formation at steadyꢀstate at 300 ºC over Ru/Al O catalysts
4
2
3
2
as a function of Ru loading (open symbols: data obtained form a 0.1% Ru/Al O3
catalyst that was tested previously at 350 ºC). (b) CO slectivity as a function of Ru
loading at 300 ºC.
Figure 6. Temperature programmed CO and CO conversions over a 0.1 % Ru/Al O catalyst.
2
2
3
Figure 1(a).
1
3
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
60
0.1%Ru/Al O
2
3
3
3
3
3
0
.5%Ru/Al O
2
5
4
3
2
1
0
0
0
0
0
0
1%Ru/Al O
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
5
%Ru/Al O
2
%Ru/Al O
2
0
100
200
300
400
500
o
Temperature( C)
1
4
ACS Paragon Plus Environment

Page 15 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
Figure 1(b)
4
0
0
0
.1%Ru/Al O
2
3
3
3
3
3
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
.5%Ru/Al O
2
1
2
5
%Ru/Al O
2
%Ru/Al O
2
30
%Ru/Al O
2
2
1
0
0
0
0
100
200
300
400
500
o
Temperature( C)
1
5
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 1(c)
6
5
4
3
0
0
.1%Ru/Al O
2
3
3
3
3
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
0.5%Ru/Al O
2
0
0
0
1
2
5
%Ru/Al O
2
%Ru/Al O
2
%Ru/Al O
2
20
1
0
0
0
100
200
300
400
500
o
Temperature( C)
1
6
ACS Paragon Plus Environment

Page 17 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
Figure 2(a)
0
.6
0
.1%Ru/Al O3
2
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
CO2
3
50oC
0.5
0.4
0.3
0.2
CH4
CO
0
.1
0
0
200
400
600
800
1000
1200
Time-on-stream/min
1
7
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 2 (b)
0
.2
0.1%Ru/Al O
at 300 C
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
o
0
0
0
0
.16
.12
.08
.04
0
x4.5
x20.7
x1.5
CO2
CO
CH4
fresh
after reaction
1
8
ACS Paragon Plus Environment

Page 19 of 26
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
Figure 3.
(
a)
(b)
19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4(a)
4
Ru/Al O
2
3
CO2
0.1%
0
1
2
5
.5%
%
%
3
2
1
0
%
1
.9
1.8
1.7
000/T(K)
1.6
1.5
1
2
0
ACS Paragon Plus Environment

Page 21 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
Figure 4(b).
6
Ru/Al O
2
3
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
CO
5
4
3
2
1
0
0
0
1
2
5
.1%
.5%
%
%
%
1
.9
1.8
1.7
000/T(K)
1.6
1.5
1
2
1
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 4(c)
6
Ru/Al O3
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CH4
0.1%
.5%
1%
5
4
3
2
1
0
0
2
5
%
%
1
.9
1.8
1.7
000/T(K)
1.6
1.5
1
2
2
ACS Paragon Plus Environment

Page 23 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
Figure 5(a)
40
30
20
10
0
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
CH4
CO
0
1
2
3
4
5
6
Ru loading(wt%)
2
3
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 5(b)
1
00
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
6
4
0
0
0
20
0
0
1
2
3
4
5
6
Ru loading(%)
2
4
ACS Paragon Plus Environment

Page 25 of 26
ACS Catalysis
1
2
3
4
5
6
7
8
9
Figure 6
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
25
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 26
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
TOC:
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
26
ACS Paragon Plus Environment