Surface structure of yttrium-modified ceria catalysts and reaction pathways from ethanol to propene

Article abstract of DOI:10.1016/j.jcat.2014.04.017

The additive effect of 31 metals on activity of ceria catalysts was studied for the conversion of ethanol to propene. Yttrium gave the best results and the propene yield showed a volcano-shaped dependence on the Y amount, being maximized at 20 atom% Y to that of Ce. The catalyst was stable due to very small carbon deposition rate which was approximately 1/100 of that on the parent CeO₂ catalyst. The water addition increased the propene yield to 30%, decreased the ethene yield to 37%, and did not change the durability. The physicochemical characterization of the Y/CeO₂ catalysts indicated that the surface of active catalysts was covered with a solid solution of CeO₂ and Y₂O₃. Pulse experiments of various candidate compounds suggested that the reaction pathways were ethanol → acetaldehyde → ethyl acetate → acetic acid → acetone → 2-propanol → propene through Tishchenko reaction, ketonization, and Meerwein-Ponndorf-Verley reduction.

Full text of DOI:10.1016/j.jcat.2014.04.017

Journal of Catalysis 316 (2014) 112–120
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Surface structure of yttrium-modified ceria catalysts and reaction
pathways from ethanol to propene
⇑
Fumitaka Hayashi, Masashi Tanaka, Dongmei Lin, Masakazu Iwamoto
Chemical Resources Laboratory, Tokyo Institute of Technology, 4259-R1-5 Nagatsuta, Midori-ku, Yokohama 226-8503, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The additive effect of 31 metals on activity of ceria catalysts was studied for the conversion of ethanol to
propene. Yttrium gave the best results and the propene yield showed a volcano-shaped dependence on
the Y amount, being maximized at 20 atom% Y to that of Ce. The catalyst was stable due to very small
carbon deposition rate which was approximately 1/100 of that on the parent CeO2 catalyst. The water
addition increased the propene yield to 30%, decreased the ethene yield to 37%, and did not change
the durability. The physicochemical characterization of the Y/CeO2 catalysts indicated that the surface
of active catalysts was covered with a solid solution of CeO2 and Y2O3. Pulse experiments of various
candidate compounds suggested that the reaction pathways were ethanol ? acetaldehyde ? ethyl
acetate ? acetic acid ? acetone ? 2-propanol ? propene through Tishchenko reaction, ketonization,
and Meerwein–Ponndorf–Verley reduction.
Received 29 December 2013
Revised 9 April 2014
Accepted 27 April 2014
Keywords:
Ceria
Yttrium
Bioethanol
Propene
Ó 2014 Elsevier Inc. All rights reserved.
Heterogeneous catalysis
1. Introduction
concentration of ethanol is essential to reduce running costs of
the industrial processes). Therefore, new catalysts workable under
Propene is one of the most important ingredients in petroleum
chemistry and its needs continuously increase because of growing
production of propene derivatives, such as polypropene, propene
oxide, and acetonitrile [1]. Today, propene is produced through
hydrothermal cracking of naphtha. Depletion in fossil fuel and
increased concerns on the environment stimulate replacing petro-
leum with renewable resources. Several researchers including us
have suggested that bioethanol is a promising renewable platform
chemical to yield various C3 and C4 compounds [2–13]. The
present study was thus devoted to the conversion of ethanol to
propene.
Zeolite catalysts are well known to show 20–30% of propene
yield, except for short-term high yields, but this selectivity
decreased with duration [14–17]. Oligomerization, cracking, and
aromatization proceed on the acid sites of the zeolite pore, and
propene is selectively produced as a result of shape selectivity of
the pore window. Such uncontrolled reactions, however, result in
coke formation and short lifetimes of catalysts. On the other hand,
Ni-ion-loaded MCM-41 [8] and Sc-modified In2O3 catalysts [9,10]
offered new types of catalysis for the production of propene
without shape selectivity, but the catalytic activity gradually
decreased with duration at 30 vol% of ethanol (the high
the practical conditions are needed to produce propene from
bioethanol.
Various metal oxide-based catalysts have been examined as
catalysts for the conversion of ethanol, but most of the products
besides ethene are oxygenated compounds, such as aldehydes,
ketones, and higher alcohols [2–7,18–21]. Our attention was paid
to the selective conversion of Cx-alcohol to C2xꢀ1-ketone on ceria-
based catalysts [2–6] and assumed that any acetone formed could
be hydrogenated and subsequently dehydrated by controlling the
acid-base and redox properties of catalysts. Indeed, yttrium-loaded
CeO2 was found by us to show the stable catalytic activity for the
selective conversion of ethanol to propene [12]. In this study, we
first examined the additive effect of 31 metals on the ceria cata-
lysts in the absence and presence of water. Niobium and yttrium
were then selected as effective additives and the dependences of
activity on their added amounts were studied. Second, effect of
the preparation method was examined for the better catalyst, Y/
CeO2. The structural, redox, and acid-base properties of Y/CeO2
were characterized to reveal the correlation between the catalytic
activity and the properties. Last, the reaction pathways were dis-
cussed on the basis of the results of product distributions in the
continuous flow and pulse reactions.
⇑
Corresponding author. Fax: +81 3 3817 1612.
E-mail address: iwamotom@tamacc.chuo-u.ac.jp (M. Iwamoto).
http://dx.doi.org/10.1016/j.jcat.2014.04.017
0021-9517/Ó 2014 Elsevier Inc. All rights reserved.

F. Hayashi et al. / Journal of Catalysis 316 (2014) 112–120
113
vacuum. Field-emission scanning electron microscopy, FE-SEM,
images were collected with a Hitachi HR-S5500. Powder X-ray
diffraction, PXRD, patterns were measured using a Rigaku Ultima
2. Experimental
2.1. Preparation of metal-modified ceria catalysts
IV diffractometer with monochromated Cu
K
a
radiation
(k = 0.15418 nm, 40 kV, 40 mA). Temperature-programmed
reduction with hydrogen, H2-TPR, profiles were recorded with a
BEL Japan BelCat. X-ray photoelectron (XP) spectra were measured
with an Omicron EA125 XPS system. All binding energies were
corrected using the values of C 1s (285.0 eV). X-ray absorption fine
structure, XAFS, analysis was carried out at the BL-9C station of the
Photon Factory at the High Energy Accelerator Research Organiza-
tion, Japan. The energy and current of electrons in the storage ring
were 2.5 GeV and 350 mA, respectively. Energy calibration was
performed using a Cu foil (8.9788 keV). Prior to the measurements,
the samples were evacuated at 673 K. Data analysis was performed
using the REX2000 program (Ver. 2.5.9, Rigaku). The k3-weighted
extended X-ray absorption fine structure, EXAFS, oscillation was
Fourier-transformed (FT) into the r space using the k range of
3.0–14.0 Åꢀ1 for Y or 3.0–10.0 Åꢀ1 for Ce. Curve fitting analysis
was carried out in the r range 1.2–2.2 Å for the Y–O contribution
or 1.5–2.3 Å for the Ce–O to calculate the average nearest neighbor
coordination number and the average bond lengths of Y–O and
Ce–O. Backscattering amplitudes and phase-shift functions were
calculated from the crystal structures of Y2O3 and CeO2 using the
FEFF6 code.
Diffuse reflectance infrared Fourier transform (DRIFT) spectra
were measured in a DR-600Bi diffuse reflectance cell using Jasco
FTIR-6300 spectrometer equipped with a MCT detector. Self-
supported KBr plates were used as windows for the cells. Scan
times and a resolution were 32 and 4 cmꢀl, respectively. The
KBr powder evacuated at 303 K was used as a reference. Prior
to the measurements, the sample was evacuated at 673 K for
1 h. Pyridine (>99.5%, Wako) or CO2 (>99%, GL Sciences) was
introduced onto the sample at 303 K for 5 min and then
evacuated at the same temperature for 30 min. All spectra were
recorded as the difference spectra before and after the introduc-
tion of adsorbates.
Ceria-based catalysts were prepared using
a
conventional
impregnation method or a co-precipitation method. In the former,
the parent ceria, obtained from Catalysis Society of Japan
(JRC-CEO-3, 85 m2 gꢀ1), was impregnated using aqueous solutions
of the metal salts: nitrates of Li, Mg, Ca, Y, La, Sm, Er, Mn, Fe, Co, Ni,
Cu, Zn, Rh, Cd, Al, In, and Bi, acetates of Sc and Sn, ammonium salts
of Ti, V, Nb, Mo, W, and Re, alkoxides of Ta and Si, and chloride of Ir.
In the latter, Ce(NO3)3ꢁ6H2O, Y(NO3)3ꢁ6H2O, and ammonium
hydrogen carbonate were used as precursors and a precipitant,
respectively, and the solid solution was prepared according to
the literature [22]. All samples were calcined at 873 K for 5 h in
air. Catalysts prepared by the impregnation and the co-precipitation
method are denoted by M(x)/CeO2 and M(x)-CeO2, respectively,
where M and x represent the metal used and the atomic % to that
of Ce.
2.2. Measurement of catalytic activity
Ethanol (>99.5%, Kanto chemical) was used without further
purification. The continuous flow reactions were conducted in a
fixed-bed plug-flow reactor made of quartz (i.d. 10 mm) at an
atmospheric pressure under the following conditions: catalyst
weight 0.05–2.0 g (particle diameter, 300–600 lm), total flow rate
12–32 ml minꢀ1, partial pressure of ethanol (PEtOH) 30 vol%, PH2O
0–30 vol%, N2 balance. Before the catalytic runs, the catalysts were
heated in a N2 flow at 673 K for 1 h. The products were analyzed
with an online automatic gas chromatograph (AG-1, Round Sci-
ence) equipped with four kinds of packed columns: (1) activated
charcoal, (2) MS 5A, (3) Porapak Q, and (4) PEG-20 M. The columns
1–3 were independently connected to the three TCD detectors to
determine the yields of ethene, CO, and CO2. The column 4 was
connected to the FID detector to determine the yields of other car-
bon products. The C4–C7 hydrocarbons and oxygenated products
were analyzed as needed with an offline FID gas chromatograph
(GC-2014, Shimadzu) with a capillary column of CP-7561 (10 m,
Varian) or GS-Q (30 m, Agilent). Conversion levels of ethanol and
yields of carbon products (CnHxOy, n = carbon number) were calcu-
lated by Eqs. (1) and (2), where the partial pressures of ethanol
3. Results and discussion
3.1. Additive effect of metals on activity of ceria catalysts
before and after the reaction are denoted by PEtOH and PE0 tOH
respectively.
,
Table 1 summarizes the activity of various ceria catalysts mod-
ified with mainly 10 atom% metal additives, in which the data after
0.75 h are shown because the activity was often changed with
duration. The metals were listed in the sequence of the family.
The surface areas of M(10)/CeO2 are also summarized in the table
and were roughly constant at 52–61 m2 gꢀ1 independent of the
metal additives.
The conversion degrees of ethanol were all 80–100% under the
present reaction conditions. Among 31 metals, the addition of Y,
Sm, Ti, Nb, or Ta increased the propene yield, while that of Fe,
Co, Cu, Zn, Sn, or Bi increased the acetone yield. The formation of
acetone on the latter group was well consistent with the previous
reports on ZnO- and CeO2-based catalysts [2–6] that were active
for the conversions of Cx-alcohols to C2xꢀ1 ketones. As shown in
Run 18 of Table 1, the Nb(10)/CeO2 catalyst exhibited the highest
propene yield of 32%, but its activity was not stable. The deactiva-
tion with duration was also observed on the Ti- or Ta-modified
CeO2 catalyst. In contrast, the Y(10)/CeO2 catalyst showed no deac-
tivation during continuous service of 80 h, although the propene
yield, 19% (Run 6), was lower than those on the Nb(10)/CeO2 and
the Ti(10)/CeO2. Based on the high propene yield or the long life,
the Nb- and Y-modified ceria were selected as the catalysts in
the subsequent detailed study.
Ethanol conv: ð%Þ ¼ 100ð1 ꢀ ðP0EtOH=PEtOHÞðPN2=PN0 2ÞÞ
ð1Þ
Cn Yield ð%Þ ¼ 100ðn P0 =2PEtOHÞðPN2=P0N2Þ
ð2Þ
Cn
For pulse experiments, ethyl acetate (>99.5%, Kanto chemical),
acetic acid (>99.5%, Kanto chemical), acetone (>99.9%, Aldrich),
and ethanol (>99.5%, Kanto chemical) were used as substrates
without further purification. Each substrate with volume of 2
was fed into the catalyst bed in the quartz glass reactor (i.d.
ll
7.6 mm) at 403 K under the following conditions: catalyst weight
0.1 g (particle diameter, 300–600 lm), total flow rate 10-
80 ml minꢀ1, N2 balance. As needed, ethanol, water, or H2 gas
was used as a co-reactant. The pretreatment and the product anal-
ysis methods were the same as those for the continuous flow
reactions.
2.3. Characterization
N2 adsorption isotherms were determined at 77 K with an
automatic gas sorption meter BEL Japan Belsorp mini II. Before
the measurements, the samples were degassed at 423 K for 2 h in

114
F. Hayashi et al. / Journal of Catalysis 316 (2014) 112–120
Table 1
Activity of various metal-modified ceria catalysts for the conversion of ethanol.a
c
Run
Additive (atom%)b
SBET
React. cond.
Temp.(K)
Conv. (%)
EtOH
Yield of products on carbon basis (%)
AC3/COx
(m2 gꢀ1
)
PH2O(vol%)
C3H6
Acetone
C2H4
C4H8
CO
CO2
Carbon Total
1
2
3
4
5
6
7
8
9
None
Li(10)
Mg(10)
Ca(10)
Sc(10)
Y(10)
Y(20)
Y(20)
Y(50)
Y(25)Solid-Sol
Y(67)Solid-Sol
Y(150)Solid-Sol
La(10)
Sm(10)
Er(10)
Ti(10)
85
693
723
723
723
723
723
703
703
723
698
698
698
723
425
698
673
723
673
698
673
698
723
723
723
723
673
673
673
673
673
723
673
723
698
673
723
723
673
673
723
0
0
0
0
0
0
0
30
0
0
0
0
0
0
0
0
0
0
30
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
85
12
1
13
9
10
14
5
7
3
0
0
2
0
12
6
6
4
6
3
0
0
1
12
0
2
9
3
11
6
36
27
9
15
16
11
9
13
12
5
31
8
2
1
6
1
0
1
1
0
1
0
0
1
0
0
3
12
8
10
4
7
8
5
19
10
0
2
0
1
1
4
6
1
3
3
1
4
5
6
4
3
3
2
3
2
3
3
3
4
3
3
2
1
1
3
1
2
5
1
1
6
5
4
6
1
2
8
12
4
3
1
7
5
2
2
1
9
5
13
9
8
12
8
9
5
6
7
78
38
83
56
55
79
92
86
71
65
76
75
70
81
58
77
72
82
81
60
71
51
80
79
49
79
52
57
37
40
49
70
48
54
33
62
62
46
41
61
1.7
1.6
0.9
1.3
1.1
1.2
2.5
2.7
2.2
1.8
2.5
2.7
1.2
2.1
0.9
1.9
0.9
3.0
2.2
1.4
2.1
0.8
1.9
2.0
0.6
2.2
1.7
0.6
1.2
1.2
0.8
0.4
1.1
1.3
1.1
0.8
1.1
1.6
1.1
4.6
n.m.d
n.m.d
57
100
100
100
100
100
100
>99
84
100
100
100
100
>99
100
100
99
100
99
100
100
98
100
100
100
100
99
29
17
24
37
52
37
45
25
36
36
33
42
21
15
8
n.m.d
59
54
54
45
9
19
25
30
15
11
20
20
11
16
10
27
9
32
32
7
26
4
7
9
5
3
2
2
1
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
46
60
48
7
9
8
n.m.d
n.m.d
n.m.d
61
11
13
9
10
17
5
12
11
4
V(10)
n.m.d
54
Nb(10)
Nb(10)
Nb(20)
Ta(10)
Cr(10)
Mo(10)
W(10)
Mn(10)
Fe(20)
Co(10)
Ni(10)
Cu(10)
Zn(10)
Re(1)
17
7
54
52
32
15
11
28
8
13
7
4
0
3
3
2
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
n.m.d
52
9
12
13
14
13
13
13
12
15
12
10
7
14
12
13
14
2
96
98
100
100
>99
>99
92
100
99
>99
82
5
2
3
6
5
5
15
1
2
Rh(1)
Ir(1)
Cd(10)
Al(10)
In(10)
0
3
15
8
13
15
3
11
4
23
15
4
7
4
2
1
Si(10)
Sn(10)
Bi(10)
5
46
e
Y2O3
10
79
7
0
a
b
c
Catalyst weight, 2.0 g; total flow rate, 12.8 ml/min; PEtOH, 30 vol% (N2 balance); total pressure, 0.1 MPa. All of the results were obtained after 45 min.
Additives were loaded using the impregnation or co-precipitation method. Subscript Solid-Sol represents the latter method, while no subscript symbol does the former.
Ratio of total yield of propene and acetone to that of CO2 and CO.
Not measured.
C-type Y2O3 used as a catalyst for comparison.
d
e
The results on Nb addition will be introduced first. The effect of
Fig. 1b shows the effect of the amount of Y. The dependency was
again of volcano shape, namely the yield of propene was maxi-
mized at 20 atom% Y. Fig. 2c and d show the long-term stability
of the Y(20)/CeO2 catalyst in the absence and presence of water,
while Fig. S1b shows the product distribution as a function of
the amount of Nb on the catalytic activity is shown in Fig. 1a and in
Table 1, Runs 18 and 20. The surface areas of the catalysts were
decreased through the Nb addition from 85 to 52–54 m2 gꢀ1 irre-
spective of the Nb amount. The propene yield was maximized at
10 atom% Nb, and it reached 32% as shown in Fig. 1a. The other
byproducts were ethene, acetone, and butenes. Fig. 2a shows the
stability of the Nb(10)/CeO2 catalyst in the absence of water. The
propene yield decreased with duration, and the acetone yield
increased. The effect of water addition was further examined to
prevent the deactivation. Fig. 2b shows the long-term experiment
at PH2O = 30 vol%, and Fig. S1a (Supplementary material) shows the
product distribution as a function of PH2O. The yield of propene
became constant in the presence of water within the present
experiment, but the yields of ethene and acetone changed gradu-
ally with the reaction time. As will be shown later, since acetone
was an intermediate for the propene formation, the increase in
the acetone yield indicated partial deactivation of the Nb(10)/
CeO2 catalyst. The detailed study on the Nb-modified catalyst
was not continued here except for the measurement of carbon
deposition.
PH2O. The yields of propene were constant at 25% and 30% during
continuous experiments for 50–80 h, and the ethene yields were
also constant at 50% and 37%, respectively. The findings indicated
the high durability of Y/CeO2 and the effectiveness of the water
addition for the increase in the propene formation and decrease
in the ethene formation. It is widely accepted that the dehydration
of ethanol is well catalyzed on acid sites. The acidity on the Y/CeO2
catalyst might be weakened by the addition of water [16]. The tol-
erance against water vapor is an advantage of the Y/CeO2 catalyst
over the zeolite catalysts, since the dealumination and the subse-
quent deactivation of zeolites are the significant problems.
The long-term stabilities of the catalysts were discussed from
the viewpoint of carbon deposition. The thermogravimetric analy-
ses gave the following carbon deposition rates (see Fig. ꢀ31): in the
absence of water, 1.8, 2.6, 0.02, and 1.7–4.7 mgcarbon gcat hꢀ1 on
the pure CeO2, Nb(10)/CeO2, Y(20)/CeO2, and H-ZSM-5 zeolites

F. Hayashi et al. / Journal of Catalysis 316 (2014) 112–120
115
50
40
30
20
10
50
40
30
20
10
0
(a)
(a)
0
0
4
8
12
16
20
Time on stream / h
0
10
20
30
Amount of Nb loaded / atom %
50
40
30
20
10
0
(b)
50
(b)
40
30
20
10
0
63 64 65 66 67 68
5
6
7
8
9
Time on stream / h
0
10
20
30
40
50
Amount of Y loaded / atom %
50
40
30
20
10
0
50
40
30
20
10
0
(c)
(c)
7
8
9
10 77 78 79 80 81 82
Time on stream / h
0
30
60
90 120 150
50
40
30
20
10
0
Amount of Y loaded / atom %
Fig. 1. Catalytic activity of Nb/CeO2 (a), Y/CeO2 (b), and Y-CeO2 (c) as a function of
the amount of Nb or Y loaded for the production of propene (red triangle), ethene
(green circle), butenes (orange square), acetone (blue triangle), and acetaldehyde
(black circle). The results obtained after 45 min are plotted. Reaction conditions:
catalyst weight, 2.0 g; total flow rate, 12.8 ml minꢀ1; PEtOH, 30 vol%; and reaction
temperature, 693–723 K. (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article.)
(d)
51 52 53 54 55 56
11 12 13 14
Time on stream / h
[14,23] and in the presence of water (PH2O = 30 vol%) 0.6, 0.4, 0.06,
Fig. 2. Long-term experiments for the catalytic activity of Nb(10)/CeO2 (a,b) and
Y(20)/CeO2 (c,d) in the absence and the presence of water. Symbols are the same as
those in Fig. 1. Reaction conditions: catalyst weight, 2.0 g; total flow rate,
12.8 ml minꢀ1; PEtOH, 30 vol%; PH2O, 0 (a,c) and 30 vol% (b,d); reaction temperature,
698 (a,b), 693 (c), and 703 K (d).
ꢀ1
and 0.6–1.6 mg
g
hꢀ1, respectively. In the absence of water,
carbon cat
the carbon deposition rate on the Y(20)/CeO2 catalyst was approx-
imately 1/100 of those on the Nb(10)/CeO2 and zeolites. In the
presence of water, most of the values were reduced, but that on
the Y(20)/CeO2 was still smaller than those on the Nb(10)/CeO2
and zeolites. It was indicated that the Y(20)/CeO2 would be the
best catalyst for the stable conversion of ethanol to propene among
the ceria-based catalysts.
Fig. 1c. The propene yields on the Y(67)– and Y(150)–CeO2 were
both 21%, which was similar to 25% on the Y(20)/CeO2. It would
be worthy to note that the propene yield on pure Y2O3 was only
7% (Run 40, Table 1).
Fig. S2 shows the FE-SEM images of the CeO2, Y(20)/CeO2, and
Y(67)–CeO2. The parent CeO2 was aggregates of nanoparticles with
10–20 nm diameters (Figs. S2a and S2b). The morphology of the
Y(20)/CeO2 was almost the same as that of CeO2 (Figs. S2c vs.
S2b), indicating no change with loading treatment of Y. The
Y(67)–CeO2 catalyst also showed the similar particle size and
morphology to those of Y(20)/CeO2 (Fig. S2d). Fig. S3 shows the
PXRD patterns of the CeO2, Y(20)/CeO2, and Y(67)–CeO2. All
3.2. Correlation of catalytic activity with physicochemical properties of
Y/CeO2
The co-precipitation of Y2O3 and CeO2 (=Y-CeO2) gave their
solid solutions, which were employed here as reference catalysts.
Their typical physicochemical properties and catalytic activity
were briefly introduced. The surface areas of the Y-CeO2 were
46–60 m2 gꢀ1 (Runs 10–12, Table 1), being almost the same as
those of the Y/CeO2 (Runs 6–9). The activity is summarized in

116
F. Hayashi et al. / Journal of Catalysis 316 (2014) 112–120
here studied by the XAFS analysis. Fig. 4a shows the Y-K-edge
ZSM-5
XANES spectra of the Y/ and Y-CeO2. Two noticeable features were
observed. First, all samples show no pre-edge peaks around
17030 eV, assignable to the 1s ? 4d transition of Y. This s ? d
transition is partially allowed for the distortion of octahedral envi-
ronment, only when p-orbitals are mixed with d orbitals. No
appearance of this transition indicated little distortion of the octa-
hedral symmetry, as has been reported in the other Y-containing
systems [26]. The second point was change in the white line shape
with the sample. All samples exhibited two peaks at 17050 and
17058 eV attributable to two inequivalent Y sites in the C-type
bixbyite structure [27], although the peak shapes were different.
The Y(10)/, Y(20)/, and Y(67)–CeO2 samples gave almost the same
spectra, indicating the equivalent distributions of Y in the fluorite
structure. On the other hand, the C-type sesquioxide Y2O3 and
heavily Y-doped Y(50)/CeO2 exhibited the distinct increments in
the 17058 eV. The change suggested the increase in the Y2O3 phase
on these samples, which will be discussed in the following para-
graphs in more detail.
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0
10 20 30 40 50
Amount of Y or Nb / atom %
Fig. 3. Carbon deposition rate on Nb/CeO2 (black symbols) and Y/CeO2 (red
symbols) as a function of the amount of Nb or Y loaded, together with that on the
zeolites (blue symbols in the right panel) reported in Refs. [14,23]. Partial pressure
of water: closed black, red, and blue circles, 0 vol%; open black and red squares,
30 vol%, blue square, 88 vol%. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)
The FT-EXAFS spectra shown in Fig. 4b gave valuable informa-
tion on the local structure. The Y(50)/CeO2 and C-type Y2O3
samples showed three bands: The first peak at 1.8 Å was due to
the Y–O shell, and the second and third peaks at 3.1 and 3.8 Å were
due to two Y–Y shells distributed at different distance [26,27]. In
contrast, the Y(10)/, Y(20)/, and Y(67)–CeO2 samples showed two
significant signals at 1.8 and 3.6 Å, which were attributed to the
Y–O and Y–O–Ce shells, respectively. The Y(20)/CeO2 sample also
gave a small peak at 3.1 Å due to the Y–O–Y shell. The appearance
of the 3.6 Å peaks for the Y(10)/, Y(20)/, and Y(67)–CeO2 was inter-
preted as the evidence for the formation of solid solution of CeO2
and Y2O3, and the disappearance of this peak for the Y(50)/CeO2
was attributed to the formation of thick Y2O3 layers on the surface.
The results of curve fitting analysis for Y species are summarized in
Table S2. The coordination number, CN, for the solid solution
samples exhibited single-phase diffraction patterns assignable to
the cubic fluorite structure of ceria (JCPDS No. 34-394,
a = 0.5410 nm). The peak positions were little shifted upon the Y
loading, indicating the high dispersion of Y ions in the ceria lattice.
The FE-SEM and PXRD analyses indicated no noticeable difference
between the Y/ and Y-CeO2 samples.
It was reported that incorporation of Y3+ cations to tetravalent
host ceria readily proceeded as a result of the similar ionic radius:
The ionic radii of Ce4+ and Y3+ ions are 0.97 Å for the 8 coordination
and 1.02 Å for the 6 coordination, respectively [24]. Such replace-
ment by Y cations led to the formation of both oxygen vacancies
and reduced Ce3+ ions [25]. The local structures of Y and Ce were
1.6
(a)
(b)
30
1.4
Y2O3
1.2
1.0
25
Y(50)/CeO2
20
0.8
0.6
0.4
0.2
0.0
17050 17058
15
10
5
Y(20)/CeO2
Y(10)/CeO2
17070
x1
Y(67)-CeO2
0
17020 17030 17040 17050 17060 17070
1
2
3
4
5
6
Photon energy / eV
Distance / angstrorm
2.5
AB
(c)
A
40
30
20
10
0
1173
1073
973
873
773
673
573
473
2.0
1.5
1.0
0.5
0.0
(d)
C
D
5710 5720 5730 5740 5750 5760
Photon energy / eV
0
2000
4000
Time / s
6000
8000
Fig. 4. Y–K-edge XANES (a) and EXAFS spectra (b) of Y(10)/CeO2 (green), Y(20)/CeO2 (blue), Y(50)/CeO2 (black), Y(67)–CeO2 (orange), and reference Y2O3 (red). Ce-LIII-edge
XANES spectra (c) and H2-TPR profiles (d) of Y(20)/CeO2 (blue) and pure CeO2 (red). Symbols A–D in (c) are the same as those in Refs. [30,31]. (For interpretation of the
references to color in this figure legend, the reader is referred to the web version of this article.)

F. Hayashi et al. / Journal of Catalysis 316 (2014) 112–120
117
Y(67)–CeO2 was 6.0, which agreed well with the results previ-
1.0
0.8
0.6
0.4
0.2
0.0
ously-reported [25,26]. Meanwhile, the CNs for the Y(10)/CeO2,
Y(20)/CeO2, and Y(50)/CeO2 were partly smaller (5.6–5.7),
indicating the formation of coordinatively unsaturated Y species.
The average Y–O bond lengths for the respective samples were
ca. 2.25 Å, independent of the catalyst preparation method and
amount of Y loading. The results could suggest the formation of
the solid solution at low loading of Y and the increase in the
Y2O3 phase at the high loading, although their amounts depended
on the catalyst preparation methods.
L
(a)
B
CeO2
Y(20)
/CeO2
Y2O3
Fig. 4c shows Ce-LIII-edge XANES spectra of the Y(20)/CeO2 and
pure CeO2 samples. Pure CeO2 was reported to usually possess
oxygen vacancies of 1 ꢂ 1012 – 1 ꢂ 1014 cm–2, whose amounts
were dependent on the crystallite size [28,29]. The spectra showed
four different lines named A, B, C, and D according to the literature
[30,31]. The observed four lines were all characteristic of a Ce4+ ion,
originating from the hybridization of Ce 4f and O 2p states in the
initial state. It should be noted, however, that certain amounts of
Ce3+ species might exist, because the peak position of Ce3+ was
very close to that of the line C of Ce4+ ion (the difference was only
1.9 eV) [30,31]. The curve fitting analysis was thus conducted to
roughly quantify the contents of Ce3+ ions and the results are sum-
marized in Table S3. The average CN for CeO2 and Y(20)/CeO2 was
7.8 and 7.4, which were lower than 8.0 for oxygen vacancy-free
ceria. When all decrement in the CN value was attributable to
the formation of oxygen vacancies accompanied by the change of
Ce4+ to Ce3+, approximately 30% of Ce was estimated to be reduced
to Ce3+ in the Y(20)/CeO2 [32].
Fig. 4d shows the TPR profiles of pure CeO2 and Y(20)/CeO2 to
determine amounts of the oxygen vacancies. The pure CeO2 exhib-
ited three reduction peaks at 790, 950, and 1170 K. The first and
second peaks were due to the reduction of surface oxygen ions,
and the third one was due to the reduction of oxygen ions in the
bulk [33–35]. The total amount of H2 consumption on the CeO2
was 2.63 mmolH2 gꢀ1 based on the integrꢀa1ted peak areas, which
was 90% of the amount (2.90 mmolH2 g ) required to reduce
defect-free ceria, CeO2 + 1/2H2 ? CeO1.5 + 1/2H2O. The observed
difference of 10% could be attributed to oxygen vacancies around
the Ce3+ ions, as has been suggested in the literature [28,29] and
is shown in Table S3. The Y(20)/CeO2 gave smaller first and second
peaks and roughly the same third peak. The total amount of H2
consumption observed was 2.26 mmolH2 gꢀ1. The H2 consumption
1700
1600
1500
1400
Wavenumber / cm-1
2.0
1.5
1.0
0.5
0.0
(b)
CeO2
Y(20)/CeO2
Y2O3
1200
1800
1600
1400
Wavenumber / cm-1
Fig. 5. Difference DRIFT spectra of pyridine (a) and CO2 (b) adsorbed CeO2, Y(20)/
CeO2, and Y2O3 after evacuation at 303 K. Capital letters L and B in (a) stand for the
Lewis and Brönsted acid sites.
Brönsted acid sites, and the others on the Lewis acid sites or hydro-
xyl groups [38]. In contrast, the Y(20)/CeO2 and Y2O3 gave very
small peaks, indicating very few acid sites on their surfaces. This
result indicated that the added Y2O3 diminished the strong acid
sites on the ceria surface. Fig. 5b shows the CO2-adsorbed DRIFT
spectra. The assignment of CO2 species is summarized in Fig. S4
[39–41]. In the five bands observed on the CeO2, the bands at
1220, 1390, 1475, and 1590 cmꢀ1 were assigned to hydrogen car-
bonate, while those at 1300 and 1590 cmꢀ1 to bidentate carbonate.
On the Y(20)/CeO2, except for the 1220 cmꢀ1 band, all bands were
blueshifted by ca. 25 cmꢀ1, and the spectrum turned out to be sim-
ilar to that on Y2O3, indicating the formation of new basic sites.
Both the pyridine and CO2 adsorption experiments indicated that
the addition of 20 atom% Y resulted in the formation of yttrium
oxide-predominating surface.
Finally, the structures of Y-loaded ceria prepared by the impreg-
nation method are shown schematically in Fig. 6 as a function of Y
loading. Below 10 atom% Y, Y atoms were used for the formation of
surface solid solution of CeO2–Y2O3. At around 20 atom% Y, the sur-
face and the inside of Y/CeO2 were nearly covered with the solid
solution. Too much Y loading (around 50 atom% Y) would cause
the formation of thick Y2O3 layers. The requisite amount of Y2O3
for monolayer coverage of ceria was estimated to be 10.9 wt%catalyst
[42] according to the method in the literature [43]. The Y2O3 con-
tent in the Y(20)/CeO2 was 11.6 wt%, supporting the formation of
1–2 layers of a surface solid solution.
for Ce0.833Y0.167O1.917, which corresponds to the Y(20)/CeO2
nominally composed of 83.3% CeO2 and 16.7% YO1.5
, was
ꢀ1
estimated to be 2.57 mmol
g
from the following equation:
H2 cat
Ce0.833Y0.167O1.917 + 0.417H2 ? Ce0.833Y0.167O1.5 + 0.417H2O.
The
assumption of 10% oxygen vacancies on the sample resulted in
ꢀ1
decrease in the H2 amount to 2.31 mmolH2
g
. This value was in
cat
excellent agreement with the experimental value, 2.26. Fig. S4
shows the Ce 3d5/2 and 3d3/2 core level XP spectra of the CeO2
and Y(20)/CeO2 catalysts, and the assignment of each signal was
summarized in Table S4. Here, the Ce3+ concentrations in the two
samples were estimated according to the method in the literature
[36,37], and the results for the peak deconvolution are shown in
the figure. The Ce3+ concentrations (Ce3+/(Ce3+ + Ce4+)) of the
CeO2 and Y(20)/CeO2 samples were estimated to be 16.9% and
19.1%, respectively, being well consistent with the results for the
TPR analysis. It follows that the solid solution of CeO2 and Y2O3
was generated through the Y loading and there was the almost
same ratio of oxygen vacancies when the Y amount was not great.
Acid-base properties of the catalysts were investigated by the
DRIFT analysis using pyridine and CO2 as probe molecules.
Fig. 5a shows the pyridine-adsorbed spectra. The CeO2 exhibited
absorption peaks at 1440, 1485, 1535, 1560, 1578, and
1600 cmꢀ1, in which the 1440 cmꢀ1 band was attributable to
pyridine adsorbed on Lewis acid sites, the 1535 cmꢀ1 band on
3.3. Reaction pathways on Y-loaded CeO2
The reaction pathways were investigated based on the product
distributions in the continuous flow reactions and the pulse exper-
iments. The total yields of carbon products summarized in Table 1
were 35–90% in all experiments, greatly depending on the kinds of
metal additives. The discrepancy between the ethanol conversion

118
F. Hayashi et al. / Journal of Catalysis 316 (2014) 112–120
Fig. 6. Schematic surface structure of Y/CeO2 catalysts prepared through the impregnation method. Surface oxygen ions are omitted for clarity.
CH3CHO þ C2H5OH þ H2O ! CH3COCH3 þ CO2 þ 3H2
ð40Þ
ð5Þ
ð6Þ
ð7Þ
degrees and the product yields would be attributed to the forma-
tion of undetected products such as carboxylic acids and esters.
We indeed observed these products using an offline capillary GC
system: However, the quantitative analyses were not conducted
here because of too long intervals of the analyses.
CH3COCH3 þ H2 ! C3H6 þ H2O
2C2H5OH ! C3H6 þ CO2 þ 3H2
CO2 þ H2 ! CO þ H2O
The correlation of the product distribution with the contact
time was examined using the continuous flow reactions to under-
stand the reaction pathways yielding acetone, propene, and eth-
ene. The results on the Y(20)/CeO2 are shown in Fig. 7. The
conversion levels of ethanol increased from 5% to 100% with
increasing contact time, the selectivity of propene increased from
1% to 33%, and the selectivity of acetone decreased from 18% to
3%, indicating the possibility that acetone was an intermediate
for propene formation. On the other hand, the selectivity toward
ethene decreased from 60% to 40%. Small amounts of butenes
(not shown) were formed in the experiments. No production of
propene upon introduction of ethene onto the catalyst was also
confirmed in a separated experiment. Both findings indicated little
possibility of metathesis reaction of ethene and butenes to give
propene.
According to the reactions, the ratio of the total yield of propene
and acetone (AC3) to those of CO2 and CO (ACOx) should be 3:1,
where contribution of the reverse water–gas shift reaction (Eq.
(7)) was counted for the ACOx calculation. The calculated AC3/COx
values are listed in Table 1. The values on the catalysts active for
the propene formation were 2.5–3.0. The good consistency indi-
cated validity of the reaction pathways (3)–(7) and the estrange-
ment from 3 would be due to the side reactions or the isobutene
formation from acetone [10,11].
Three reaction sequences have been proposed for the formation
of C2xꢀ1 ketones from Cx-alcohols on oxide-based catalysts: (i)
aldol reaction [5,6], (ii) Tishchenko reaction [2,3,44], and (iii) keto-
nization reaction of carboxylic acids [45]. The possibility of the
aldol reaction will be first discussed. The aldol reaction of acetalde-
hyde gives 3-hydroxybutylaldehyde, which can be converted to
CH3C(O)CH2CHO or CH3CH = CHCHO by dehydrogenation or dehy-
dration. The latter does not give acetone, but the reaction of the
former with water might yield acetone, CO2, and H2. Since unstable
b-hydroxyaldehyde and b-ketoaldehyde could not be used as sub-
strates, the products produced in Fig. 7 were carefully analyzed
using the offline GC system. However, no production of these com-
pounds was found, suggesting little possibility of the aldol reaction
route. The formation of acetaldehyde, ethyl acetate, acetic acid, and
2-propanol was confirmed: Therefore, the possibilities of
Tishchenko reaction route (via ethyl acetate) [2,3,44] and the
ketonization route (via acetic acid) [45] were investigated in the
next section.
The reaction pathways yielding acetone from ethanol have been
suggested as Eqs. (3), (4), and (4)0 on oxide-based catalysts [2–6,9–
12]. Propene can be formed from acetone through hydrogenation
and subsequent dehydration, as described by Eq. (5) [9,10]. The
comprehensive reaction for the propene formation is thus Eq. (6).
C2H5OH ! CH3CHO þ H2
ð3Þ
ð4Þ
2CH3CHO þ H2O ! CH3COCH3 þ CO2 þ 2H2
80
60
40
20
Pulse experiments were carried out on the Y(20)/CeO2 using
ethyl acetate (AcOEt), acetic acid (AcOH), and acetone as the sub-
strates. AcOEt could be obtained from two acetaldehyde molecules
through the Tishchenko reaction [2,3,44], as described by Eq. (8).
0
100
2CH3CHO ! CH3CðOÞOCH2CH3
ð8Þ
80
60
40
20
0
Fig. 8 shows the results for AcOEt as a function of contact time. In
the absence of water, the major product was ethene irrespective
of the contact time, although the yields considerably varied from
40% to 60%. The yield of acetone was high (ꢃ20%) at the short con-
tact time of 0.11 sꢀ1, and the yields of propene were low (6–20%)
compared to those in the flow reaction system. The co-feed of water
with ethanol noticeably increased the yields of both propene and
acetone to 10–20%. The results indicated the possibility that AcOEt
was an intermediate for acetone and ethene formation.
0.0
1.0
2.0
3.0
Contact time / s
Fig. 7. Ethanol contact time dependence of the catalytic activity of Y(20)/CeO2 for
the continuous flow reactions. Reaction conditions: catalyst weight, 0.05–2.0 g;
total flow rate, 12–32 ml minꢀ1; PEtOH, 30 vol%; PH2O, 30 vol%; reaction temperature,
703 K. Symbols are the same as those in Fig. 1 except for ethanol (open black
square).
The results of AcOH are also shown in Fig. 8. AcOH would hardly
be generated from ethanol under the non-oxidative condition, but

F. Hayashi et al. / Journal of Catalysis 316 (2014) 112–120
119
AcOEt
AcOEt
+H2O
AcOH
Acetone Acetone
Ethanol
+H2
+Ethanol
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1.1 2.3 4.5 9.0 1.1 2.3 9.0 1.1 2.3 4.5 9.0 2.3
1.1 2.3 4.5 9.0
2.34.5 9.0
Fig. 8. Pulse experiment results for the Y(20)/CeO2 as a function of contact time. Reactants are ethyl acetate (AcOEt), acetic acid (AcOH), acetone, and ethanol. As needed,
water, H2 gas, ethanol was fed with the reactants, such as AcOEt and acetone. Reaction conditions: pulse size, 2.0
l, catalyst weight, 0.1 g; total flow rate, 10–80 ml minꢀ1
l
;
reaction temperature, 703 K. Red, propene; green, ethene; orange, CO2; blue, acetone; pink, AcOH; black, ethanol; gray, AcOEt; purple, acetaldehyde; white, others. (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
in the presence of water, which is a by-product of ethanol dehydra-
tion, AcOH might be yielded through the direct reaction of an
adsorbate CH3CO with a surface OH group. All introduced AcOH
was converted to acetone and CO2 even at the short contact time,
wherein the ratios of acetone produced to CO2 were almost 3:1
(carbon basis) except for that at the long contact time. These
results indicated the very easy and selective ketonization of AcOH
(Eq. (9)), agreeing well with the results reported on ceria-based
catalysts [45].
formation of oxygen vacancy [47]. The adsorbed ethanol on the
acid-base pair could reduce the acetone adsorbed on the acid site
to 2-propanaol, followed by the dehydration to give propene. The
co-fed water might be involved in this reduction, but we have
not understood yet the role of water at the molecular level.
Based on the above results and discussion, the following reac-
tion pathways yielding propene from ethanol were suggested on
the Y-modified ceria catalyst. Ethanol is dehydrogenated to acetal-
dehyde, and resulting acetaldehyde is dimerized to give ethyl ace-
tate via a Tischenko reaction. Ethyl acetate is then decomposed to
give ethene and acetic acid, the latter of which yields acetone and
CO2 through the ketonization reaction. Propene was produced by
the MPV reduction of acetone by ethanol and the subsequent
dehydration reaction. The present mechanism on Y/CeO2 is differ-
ent from that for Ni-MCM-41 [8] and Sc/In2O3 catalysts [9,10]. It
should be noted that much amount of ethene was produced at
the short contact time in the pulse reaction of ethanol (Fig. 8),
which would suggest the presence of an alternative route (Eq.
(10)) to form ethene through the simple dehydration of ethanol.
2CH3CðOÞOH ! CH3COCH3 þ CO2 þ H2O
ð9Þ
The following reaction pathways could be suggested on the basis of
the above pulse experiments, although surface intermediates
should be detected by for example IR measurements in the future.
That is, AcOEt was converted to AcOH through elimination of eth-
ene (Eq. (10)), and the subsequent self-ketonization of produced
AcOH yielded acetone and CO2 (Eq. (9)) [45].
CH3CðOÞOCH2CH3 ! CH2 ¼ CH2 þ CH3CðOÞOH
ð10Þ
Finally, the conversion of acetone to propene was confirmed to
progress on the catalyst. Differently from the reported reaction
[10], acetone gave only a small amount of propene in the presence
of H2 (propene yield was less than 10%), as shown in Fig. 8. In con-
trast, acetone could be converted to propene in the co-feed of eth-
anol, suggesting the reduction of acetone by ethanol and the
subsequent dehydration of produced 2-propanol to propene. The
former reaction is recognized as Meerwein–Ponndorf–Verley
(MPV) reduction [46], and the possible reaction scheme is shown
in Fig. 9. The active site could be the acid-base pair consisting of
Ce3+ or Y3+ and O2ꢀ ions, which was generated through the
4. Conclusions
The addition effect of 31 metal ions on ceria catalysts was
examined for the conversion of ethanol to propene. Twenty atom%
Y-loaded ceria showed a good selectivity toward propene and
remarkable durability in the absence of water. The maximum yield
of propene on the Y(20)/CeO2 was 37% at PEtOH = PH2O = 30 vol%.
The characterization revealed the formation of solid solution of
CeO2 and Y2O3. The elimination of strong acid sites originally
existing on CeO2 by the addition of Y2O3 would improve the
Fig. 9. Possible reaction mechanism for the Meerwein–Ponndorf–Verley reduction of acetone by ethanol.

120
F. Hayashi et al. / Journal of Catalysis 316 (2014) 112–120
[16] M. Inaba, K. Murata, M. Saito, I. Takahara, Green Chem. 9 (2007) 638.
[17] Z. Song, A. Takahashi, N. Mimura, T. Fujitani, Catal. Lett. 131 (2009) 364.
[18] J.I. Di Cosimo, V.K. Diez, M. Xu, E. Iglesia, C.R. Apesteguia, J. Catal. 178 (1998)
499.
durability of catalyst. The suggested reaction pathway was
ethanol ? acetaldehyde ? ethyl acetate ? acetic acid ? acetone
? 2-propanol ? propene.
[19] M.J.L. Gines, E. Iglesia, J. Catal. 176 (1998) 155.
[20] A.S. Ndou, N. Plint, N.J. Coville, Appl. Catal. A 251 (2003) 337.
[21] D.J. Elliott, F. Pennella, J. Catal. 119 (1989) 359.
Acknowledgments
[22] Y. Wang, H. Kageyama, T. Mori, H. Yoshikawa, J. Drenman, Solid State Ionics
117 (2006) 1681.
[23] R. Barthos, A. Széchenyi, F. Solymosi, J. Phys. Chem. B 110 (2006) 21816.
[24] R.D. Shanon, Acta Cryst. A32 (1976) 751.
[25] Z. Li, T. Mori, F. Ye, D. Ou, G.J. Auchterlonie, J. Zou, J. Drennan, J. Phys. Chem. C
116 (2012) 5435.
[26] P. Li, I.-W. Chen, J.E. Penner-Hahn, Phys. Rev. B 48 (1993) 10074.
[27] M. Malvestuto, R. Carboni, F. Boscherini, F. D’Acapito, S. Spiga, M. Fanciulli, A.
Dimoulas, G. Vellianitis, G. Mavrou, Phys. Rev. B 71 (2005) 075318.
[28] I. Kosacki, T. Suzuki, H.U. Anderson, H.P. Colomban, Solid State Ionics 149
(2002) 99.
[29] Y. Namai, K. Fukui, Y. Iwasawa, J. Phys. Chem. B 107 (2003) 11666.
[30] A. Martınez-Arias, M. Fernandez-Garcia, L.N. Salamanca, R.X. Valenzuela, J.C.
Conesa, J. Soria, J. Phys. Chem. B 104 (2000) 4038.
We greatly acknowledge Grants-in-Aid from the New Energy
and Industrial Technology Development Organization (NEDO)
and the Advanced Low Carbon Technology Program (JST) of Japan.
The authors appreciate Dr. Tetsuo Suzuki and Mrs. Osamu Takah-
ashi, Hiroshi Ohashi, and Takahiro Kakinuma of the NEDO research
group for helpful discussions. The measurements of the XAFS spec-
tra were performed under proposal numbers of 2009G661, and
2011G146 of the Photon Factory Program Advisory Committee.
Appendix A. Supplementary material
[31] A.V. Soldatov, T.S. Ivachenko, S. Della Longa, A. Kotani, Y. Iwamoto, A. Bichcom,
Phys. Rev. B 50 (1994) 5074.
[32] The CN for cerium is calculated to be 7.4, assuming that 70% of Ce ions are Ce4+
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2014.04.017.
(CN = 8.0) and the rest are reduced Ce3+ (CN = 6.0).
[33] V. Perrichon, A. Laachir, G. Bergeret, R. Fréty, L. Tournayan, O. Touret, J. Chem.
Soc., Faraday Trans. 90 (1994) 773.
[34] R. Körner, M. Ricken, J. Nolting, I. Riess, J. Solid State Chem. 78 (1989) 136.
[35] K. Sardar, H.Y. Playford, R.J. Darton, E.R. Barney, A.C. Hannon, D. Tompsett, J.
Fisher, R.J. Kashtiban, J. Sloan, S. Ramos, G. Cibin, R.I. Walton, Chem. Mater. 22
(2010) 6191.
[36] A. Pfau, K.D. Schierbaum, Surf. Sci. 321 (1994) 71.
[37] M. Romeo, K. Bak, J. El Fallah, F. le Normand, L. Hilaire, Surf. Interface Anal. 20
(1993) 508.
[38] M.I. Zaki, M.A. Hasan, F.A. Al-Sagheer, L. Pasupulety, Colloids Surf. A 190 (2001)
261.
References
[1] I. Chorkendorff, J.W. Niemantsverdriet, Concepts of Modern Catalysis and
Kinetics, Wiley-VCH, Weinheim, Germany, 2007. pp. 1–21.
[2] T. Nakajima, T. Yamaguchi, K. Tanabe, J. Chem. Soc. Chem. Commun. (1987)
394.
[3] T. Nakajima, K. Tanabe, T. Yamaguchi, I. Matsuzaki, S. Mishima, Appl. Catal. 52
(1989) 237.
[4] R. Sreerama-Murthy, P. Patnaik, P. Sidheswaran, M. Jayamani, J. Catal. 109
(1988) 298.
[5] Y. Kamimura, S. Sato, R. Takahashi, T. Sodesawa, T. Akashi, Appl. Catal. A 252
(2003) 399.
[6] Y. Kamimura, S. Sato, R. Takahashi, T. Sodesawa, M. Fukui, Chem. Lett. (2000)
232.
[7] T. Tsuchida, J. Kubo, T. Yoshioka, S. Sakuma, T. Takeguchi, W. Ueda, J. Catal. 259
(2008) 183.
[8] M. Iwamoto, K. Kasai, T. Haishi, ChemSusChem 4 (2011) 1055.
[9] S. Mizuno, M. Kurosawa, M. Tanaka, M. Iwamoto, Chem. Lett. 41 (2012) 892.
[10] M. Iwamoto, S. Mizuno, M. Tanaka, Chem. Eur. J. 19 (2013) 7214.
[11] J. Sun, K. Zhu, F. Gao, C. Wang, J. Liu, C.H.F. Peden, Y. Wang, J. Am. Chem. Soc.
133 (2011) 11096.
[12] F. Hayashi, M. Iwamoto, ACS Catal. 3 (2013) 14.
[13] G.R.M. Dowson, M.F. Haddow, J. Lee, R.L. Wingad, D.F. Wass, Angew. Chem. Int.
Ed. 52 (2013) 9005.
[39] C. Binet, M. Daturi, J.C. Lavalley, Catal. Today 50 (1999) 207.
[40] C. Morterra, G. Cerrato, L. Ferroni, J. Chem. Soc., Faraday Trans. 91 (1995) 125.
[41] H. Tsuneoka, K. Teramura, T. Shishido, T. Tanaka, J. Phys. Chem. C 114 (2010)
8892.
[42] The requisite amount of Y2O3 (Y2O3surf) for its monolayer coverage on the CeO2
is calculated by the following equation on the assumption that Y3+ cations are
isomorphously-substituted by Ce4+ cations on surface: Y2O3surf (wt%) = 100
FWYO1.5 SD SA 1018/Av const. Here, FWYO1.5 is the formula weight of YO1.5
,
112.9, SD is the surface density of Y atom, 6.83 Y nmꢀ2, SA is the surface area
of parent CeO2, 85 m2 gꢀ1, and Av const. is the Avogadro constant, 6.02 1023
Details are shown in Ref. [43].
.
[43] N. Naito, N. Katada, M. Niwa, J. Phys. Chem. B 103 (1999) 7206.
[44] T. Seki, T. Nakajo, M. Onaka, Chem. Lett. 35 (2006) 824.
[45] M. Renz, Eur. J. Org. Chem. (2005) 979.
[46] D. Klomp, T. Mashmeyer, U. Hanefeld, J.A. Peters, Chem. Eur. J. 10 (2004) 2088.
[47] S. Sato, F. Sato, H.i. Gotoh, Y. Yamada, ACS Catal. 3 (2013) 721.
[14] A.T. Aguayo, A.G. Gayubo, A. Atutxa, M. Olazar, J. Bilbao, Ind. Eng. Chem. Res. 41
(2002) 4216.
[15] J. Bi, M. Liu, C. Song, X. Wang, X. Guo, Appl. Catal. B 68 (2011) 107.