Condensation reactions of propanal over CexZr 1-xO2 mixed oxide catalysts

Article abstract of DOI:10.1016/j.apcata.2010.06.048

Vapor phase condensation reactions of propanal were investigated over Ce_xZr_1-xO₂ mixed oxides as a model reaction to produce gasoline range molecules from short aldehydes found in bio-oil mixtures. Several operating parameters were investigated. These included the type of carrier gas used (H₂ or He) and the incorporation of acids and water in the feed. Propanal is converted to higher carbon chain oxygenates on Ce _xZr_1-xO₂ by two pathways, aldol condensation and ketonization. The major products of these condensation reactions include 3-pentanone, 2-methyl-2-pentenal, 2-methylpentanal, 3-heptanone and 4-methyl-3-heptanone. It is proposed that the primary intermediate for the ketonization path is a surface carboxylate. The presence of acids in the feed inhibits the aldol condensation pathway by competitive adsorption that reduces the aldehyde conversion. Water also promotes ketonization and inhibits aldol condensation by increasing the concentration of surface hydroxyl groups that enhance the formation of surface carboxylates with the aldehyde. Hydrogen enhances cracking and production of light oxygenates and hydrocarbons. The light oxygenates may in turn be reincorporated into the reaction path, giving secondary products. However, the hydrocarbons do not react further. Analysis of the fresh and spent catalysts by XPS showed varying degrees of reduction of the oxide under different operating conditions that were consistent with the reaction results. Changing the proportion of the parent oxides showed that increased Zr favored formation of aldol products while increased Ce favored ketonization. This occurs by shifting the balance of the acid-base properties of the active sites.

Full text of DOI:10.1016/j.apcata.2010.06.048

Applied Catalysis A: General 385 (2010) 80–91
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Condensation reactions of propanal over CexZr1−xO2 mixed oxide catalysts
Anirudhan Gangadharan, Min Shen, Tawan Sooknoi1, Daniel E. Resasco∗, Richard G. Mallinson∗
Center for Biomass Refining and School of Chemical, Biological and Materials Engineering, University of Oklahoma, Norman, OK, 73019, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Vapor phase condensation reactions of propanal were investigated over CexZr1−xO2 mixed oxides as a
model reaction to produce gasoline range molecules from short aldehydes found in bio-oil mixtures.
Several operating parameters were investigated. These included the type of carrier gas used (H2 or He)
and the incorporation of acids and water in the feed. Propanal is converted to higher carbon chain oxy-
genates on CexZr1−xO2 by two pathways, aldol condensation and ketonization. The major products of these
condensation reactions include 3-pentanone, 2-methyl-2-pentenal, 2-methylpentanal, 3-heptanone and
4-methyl-3-heptanone. It is proposed that the primary intermediate for the ketonization path is a surface
carboxylate. The presence of acids in the feed inhibits the aldol condensation pathway by competitive
adsorption that reduces the aldehyde conversion. Water also promotes ketonization and inhibits aldol
condensation by increasing the concentration of surface hydroxyl groups that enhance the formation of
surface carboxylates with the aldehyde. Hydrogen enhances cracking and production of light oxygenates
and hydrocarbons. The light oxygenates may in turn be reincorporated into the reaction path, giving
secondary products. However, the hydrocarbons do not react further. Analysis of the fresh and spent
catalysts by XPS showed varying degrees of reduction of the oxide under different operating conditions
that were consistent with the reaction results. Changing the proportion of the parent oxides showed that
increased Zr favored formation of aldol products while increased Ce favored ketonization. This occurs by
shifting the balance of the acid–base properties of the active sites.
Received 10 March 2010
Received in revised form 22 June 2010
Accepted 25 June 2010
Available online 6 July 2010
Keywords:
Coupling
Aldol condensation
Ketonization
Biofuels
Ceria
Zirconia
Propanal
Mixed oxides
XPS
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
components as they possess reactive carbonyl moieties that can be
used to carry out carbon–carbon bond forming reactions, thereby
The production of fungible fuels from biomass offers a possi-
ble contribution to meet growing energy demands and to reduce
their environmental impact. The conversion of solid biomass can
be achieved through biochemical or thermochemical processes
[1]. Biochemical processes involve fermentation of the biomass
feedstock with a biological organism to produce ethanol or other
products. Pyrolysis is a high temperature process that produces a
liquid product that is commonly referred to as “bio-oil” [1]. Bio-oil
is a complex mixture of a wide range of oxygenated compounds
that include aldehydes, ketones, alcohols and carboxylic acids [2].
The catalytic conversion of glycerol that is obtained as a byprod-
uct from the production of biodiesel also produces short aldehydes
and acids [3]. The oxygenates found in bio-oil and from the conver-
sion of glycerol can be converted to more suitable molecular weight
moving to the transportation fuel range while reducing the oxygen
content of the oxygenate stream.
Ketonization and aldol condensation reactions are important
routes in which aldehydes (or carbonyl containing species) can
be converted into longer chain molecules. While ketonization pro-
duces higher ketones from aldehydes with the evolution of CO2 and
water, aldol condensation produces dimers and trimers of the cor-
responding aldehydes with the evolution of water. Previous studies
on ketonization of oxygenates have been carried out on a vari-
ety of metal oxide catalysts that include CeO2, ZnO, TiO2, ZrO2,
CeO2/Mn2O3, CeO2/ZrO2 and CeO2/Fe2O3 [4–13,21–23,26–32].
It has been reported that ceria-based catalysts are active for
ketonization of aldehydes, in which the aldehyde is oxidized to a
carboxylate type species on the surface that then couples to pro-
duce the ketone [11,21,22,28–30,32]. Cerium oxide is known to
possess a very high oxygen exchange capacity, due to its ability to
alternate between Ce3+ and Ce4+ under reducing and oxidizing con-
ditions, respectively [12,16–19,39]. The oxygen storage capacity of
ceria has enabled its use as an oxygen carrier, but deactivation of
ceria at high temperatures has led to the formulation of a new range
of mixed oxides that have improved stability and redox properties
[15,16]. The incorporation of ZrO2 into the lattice of CeO2 signifi-
cantly changes the reducibility and oxygen storage capacity of the
∗
Corresponding authors at: University of Oklahoma, Chemical Engineering, 100
E. Boyd Street, Room T335, Norman, OK 73019, United States. Tel.: +1 405 325 4378;
fax: +1 405 325 5813.
E-mail addresses: resasco@ou.edu (D.E. Resasco), mallinson@ou.edu
(R.G. Mallinson).
1
Permanent address: King Mongkut Institute of Technology Ladkrabang,
Bangkok, Thailand.
0926-860X/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcata.2010.06.048

A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
81
oxide [16–19]. This has been attributed to changes to the fluorite
structure of ceria and the creation of defects that increase oxygen
mobility through the lattice [16–19]. In addition, the ceria–zirconia
mixed oxides exhibit acid and base properties that can be utilized
for both aldol condensation and ketonization reactions.
In this study, the ketonization and aldol condensation reactions
of a representative short aldehyde, propanal, are carried out over a
Ce0.5Zr0.5O2 mixed oxide catalyst. The effects of different operating
conditions have been tested, including changing the type of carrier
gas used (H2 or He) and incorporating short organic acids or water
in the feed. A reaction pathway for the conversion of propanal is
proposed and related to the structural properties of the catalyst.
In addition, for comparison, changes in the ratio of the component
oxides were studied.
placed on a stainless steel holder and into an ex situ vacuum cham-
ber within the bag and then kept in the vacuum chamber for the
analysis. The XPS data from the regions related to C (1s), O (1s),
Ce (3d) and Zr (3d) were recorded for each sample. The binding
energies were corrected with reference to carbon at 284.8 eV. The
reducibility of CexZr1−xO2 catalysts were evaluated by calculating
the fraction of Ce4+ and Ce3+ in the samples by measuring the ratio
of the area of the peak at 916.5 eV to the total area of the Ce (3d)
spectrum, following a previously reported method [19,46,47].
2.4. Catalytic activity
The catalytic reactions were carried out in a 6 mm quartz tube
reactor at atmospheric pressure at 400 ◦C. Prior to the reaction, the
catalyst was reduced in situ in flowing H2 (30 ml/min) at 400 ◦C. The
feeds were propanal (from Aldrich), with, in some cases, propionic
acid (from Aldrich) that were injected from a syringe pump into
a heated port with a stream of flowing carrier gas (He or H2). To
test the effect of water addition with the feed, deionized water was
injected with another syringe pump into the reactant line. Typ-
ical catalyst loads were 15–160 mg (40–60 mesh), with the feed
and carrier flow rates adjusted to achieve the desired space time
(W/F). All the lines were heated with heating tapes to keep the reac-
tants and products in the gas phase. The products were analyzed
and identified by online gas chromatography using an HP 5890 GC
equipped with a flame ionization detector and a Shimadzu GC/MS,
respectively. Analysis of the light hydrocarbon products was done
using a Carle series 400 AGC equipped with TCD.
2. Experimental
2.1. Catalyst preparation
CexZr1−xO2 mixed oxide catalysts were prepared by a co-
precipitation method with aqueous solutions of cerium (IV)
ammonium nitrate and zirconium nitrate as described in the
studies by Hori et al. [18] and Noronha et al. [19]. Cerium (IV)
ammonium nitrate and zirconium nitrate were dissolved in water
in appropriate concentrations to have the required Ce/Zr ratio.
Then the ceria and zirconia hydroxides were co-precipitated by
increasing the pH through the controlled addition of ammonium
hydroxide. The resulting solids were washed with deionized water,
dried overnight in an oven and calcined at 773 K for 1 h. Pure CeO2
and ZrO2 were also prepared using the same method for com-
parison. ␥-Alumina was purchased from Alfa Aesar and used as
received.
2.5. Temperature-programmed desorption (TPD) of propanal
In each TPD experiment, 50 mg of fresh catalyst sample was ini-
tially pretreated in flowing H2 at 400 ◦C in the same diameter quartz
tube as used in the continuous flow experiments. After pretreat-
ment, the excess H2 was removed by purging with He. 5 ␮l (liquid)
of propanal was then repeatedly injected over the sample using
a GC syringe (Hamilton 10 ␮l) at 400 ◦C until reaching saturation,
as indicated by a steady MS signal. Excess propanal was removed
by flowing He. The sample was then cooled down to room tem-
perature in He and heated to 900 ◦C at a heating rate of 10 ◦C/min.
Masses (m/z) of 16, 18, 28, 44 and 58 were monitored continuously
by a Cirrus mass spectrometer (MKS) to determine the evolution of
methane, water, carbon monoxide, carbon dioxide and propanal,
respectively.
2.2. Catalyst characterization
The catalysts were characterized by X-ray diffraction and
temperature-programmed reduction (TPR).
The phase structure of the CexZr1−xO2 catalyst was investigated
by X-ray diffraction (XRD) using a Rigaku Automatic diffractometer
(Model D-MAX A) with a curved crystal monochromator and sys-
tem setting of 40 kV and 35 mA. Data were collected in the angle
range of 5–70◦ with a step size of 0.05◦ and a count time of 1.0 s.
The samples were finely ground and placed on a glass slide using a
smear mount technique.
TPR experiments were carried out in a gas mixture of 5% H2
in Ar (25 ml/min) over a 30 mg catalyst sample. The temperature
was increased from 303 K to 1223 K at 10 K/min. The effluent gases
were analyzed using an online SRI 110 thermal conductivity detec-
tor (TCD). Prior to each TPR experiment, the sample was pretreated
in 2% O2 in He (30 ml/min) at 550 ◦C for 1 h.
3. Results and discussion
3.1. Catalyst characterization
The textural properties of the oxides were characterized using
X-ray diffraction (XRD). It is well known from the literature that
pure CeO2 exhibits a cubic phase while pure ZrO2 shows a mix-
ture of monoclinic and tetragonal phases [16,18–20]. The distinct
peaks obtained for the mixed oxide samples in this study were
consistent with those that have been reported previously in the
literature [16,18–20] and these peaks can be assigned to contribu-
tions from both the CeO2 and ZrO2 phases. Shifts to higher angles
were observed with increasing zirconia content in the sample, and
are indicative of a change in the lattice parameter resulting from
the incorporation of the smaller Zr4+ ion into the lattice of the larger
Ce4+ ion.
2.3. X-ray photoelectron spectroscopy
X-ray photoelectron spectroscopy data were recorded on a
Physical Electronics PHI 5800 ESCA system with monochromatic
Al K␣ X-rays (1486.6 eV) operated at 100 W and 15 kV in a cham-
ber pumped down to a pressure of approximately 1.0 × 10−8 Torr.
A ∼0.64 mm2 spot size and 58.7 eV pass energy were typically used
for the analysis. The electron takeoff angle was 45◦ with respect to
the sample surface. The reduction and reaction of the samples were
carried out in a fixed bed reactor with sealing valves at the top and
bottom of the reactor. After each reaction experiment, the valves
were closed and the reactor with the sample was transferred to a
glove bag that was then purged with He to avoid any exposure to
the atmosphere. The samples were removed from the reactor and
The reducibility of the mixed oxides formed was character-
ized using temperature-programmed reduction (TPR) with H2. This
technique has been widely used to characterize the reducibility of
ceria-based materials in several previous studies and the results

82
A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
Fig. 1. Reaction of propanal as a function of W/F at 400 ◦C on Ce0.5Zr0.5O2 in He. (a) Conversion and product yield. (b) Product selectivities.
obtained here are in agreement with those [16–20]. The TPR pro-
file for pure CeO2 showed two distinct regions, a low temperature
region (300–600 ◦C) with a maximum at around 500 ◦C, attributed
to the surface reduction of the Ce4+ ions and a high temperature
region (700–900 ◦C), with a maximum at around 900 ◦C, originat-
ing from the reduction of the bulk oxide. The TPR profiles for the
mixed oxides showed a single major reduction peak centered at
temperatures between 500 and 600 ◦C. This result is indicative of
the mixed oxide character of the catalyst and is due to the defects
created by the insertion of Zr4+ into the lattice of Ce4+, as revealed
by XRD, thus increasing the oxygen mobility through the lattice. As
a result, the reduction now involves both the surface and bulk in a
single step and also occurs at a much lower temperature compared
to that of pure ceria. It is to be noted that there was no reduction of
pure ZrO2.
maximum at W/F of 0.1 and gradually decreases to a constant value
with increasing W/F. The decreases in the yields and selectivities of
3-pentanone, 2-methyl-2-pentanal and lights with increasing W/F
indicate further reactions of these compounds to yield secondary
products.
The formation of 3-pentanone on ceria-based catalysts has been
reported previously in the literature [21,22,28,29,36]. Kamimura
et al. [21,36] have proposed that the pathway to 3-pentanone
from propanal is via an aldol addition intermediate, 3-hydroxy-
2-methylpentanal, on CeO2–MgO and CeO2–Fe2O3 catalysts. They
have reported that CO and CO2 were evolved, consistent with a pro-
posed oxidative decarboxylation of 3-hydroxy-2-methylpentanal
on CeO2–Fe2O3 and deformylation on CeO2–MgO to give 3-
pentanone. They also speculated that the extra oxygen necessary
for the decarboxylation step comes from either water or an oxy-
gen impurity in the carrier gas. However, Claridge et al. [22]
have proposed that the extra oxygen in the ketonization to
form 3-pentanone from propanal comes from the surface of the
catalyst. Lietti et al. [23,26] have studied the reactions of C3 oxy-
genates (1-propanol, propanal and propanoic acid), 3-pentanone
and 2-butanone over ZnCrOx and K2O–ZnCrOx catalysts using
temperature-programmed reactions and flow experiments. A prod-
uct distribution similar to the one obtained in this study was
reported and a general reaction network was also proposed
for the transformation of oxygenates over the ZnCrOx catalyst
that included hydrogenation-dehydrogenation, aldol condensa-
tion, ketonization, dehydration, decarboxylation and cracking
reactions to yield the observed products [23,26].
Those previous studies have shown that propanal can either
oxidize to a carboxylate type species on the surface and undergo
further coupling to give 3-pentanone, or undergo aldol conden-
sation to form the hydrated dimer (3-hydroxy-2-methylpentanal)
that can decompose to give 3-pentanone. The former case requires
lattice oxygen for the propanal to form a surface propionate species
that then couples to give the ketone. Neither the hydrated dimer
nor propionic acid was observed in the gas-phase as intermedi-
ate products over the ceria–zirconia catalyst in our study. CO2
was the major gaseous product observed, while CO was seen
only in trace amounts. Also, a steady decline in catalytic activ-
ity, to around one third of the initial activity was also observed
(not shown here) over a 4 h reaction period, suggesting possible
oxygen uptake from the surface. These results suggest that the
formation of 3-pentanone over the ceria–zirconia catalyst occurs
through a decarboxylative condensation reaction via a surface
3.2. Product distribution as a function of W/F at 400 ◦C in helium
Fig. 1(a) shows the conversion and yields of products at dif-
ferent W/F from the conversion of propanal over the Ce0.5Zr0.5O2
catalyst under He flow. The products obtained include a mix-
ture of aldehydes (2-methylpropanal, 2-methyl-2-pentenal, and
2-methylpentanal), ketones (2-butanone, 3-methyl-2-butanone,
2-methyl-3-pentanone, 3-pentanone, 3-heptanone, 4-methyl-3-
heptanone, 4-methyl-3-heptenone and other higher ketones),
hydrocarbons (methane, ethane, ethylene, propane, and propy-
lene) and other light gases (CO and CO2). Here, 2-methylpropanal,
2-butanone, 3-methyl-2-butanone and the hydrocarbons are
reported as lights while 3-heptanone, 4-methyl-3-heptanone, 4-
methyl-3-heptenone and other heavier ketones are reported as
C6–C9+ coupling products.
From Fig. 1(a), it can be observed that 3-pentanone, 2-methyl-
2-pentenal and the C6–C9+ coupling products are the dominant
products from the reaction. At low conversions (W/F < 0.2), the
yields of 3-pentanone and 2-methyl-2-pentenal are comparable, as
observed from the slope of the curves. As the conversion increases,
the yields of 3-pentanone, C6–C9+ products and 2-methylpentanal
increase, while the yields of 2-methyl-2-pentenal and lights reach
a maximum and start to decrease, as shown in Fig. 1(a), suggesting
that these are primary products. This can be seen more clearly in
the plot of selectivity as a function of W/F, Fig. 1(b). A maximum
in selectivity to the lights and 2-methyl-2-pentenal is reached at
around W/F of 0.1 and with a further increase in W/F, they decrease.
It can also be seen that the selectivity to 3-pentanone reaches a

A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
83
Fig. 2. Reaction of propanal as a function of W/F at 400 ◦C on Ce0.5Zr0.5O2 in H2. (a) Conversion and product yield. (b) Product selectivities.
propionate intermediate, using lattice oxygen. The aldol condensa-
tion pathway to give 2-methyl-2-pentenal proceeds in parallel via
the aldol addition dimer, 3-hydroxy-2-methylpentanal. The pres-
ence of additional products: 4-methyl-3-heptanone (ketonization),
2-methylpropanal, 3-heptanone (aldol condensation), 2-methyl-
3-pentanone (reverse aldol condensation), 2-butanone and 3-
methyl-2-butanone (reverse ␣-addition) and light hydrocarbons,
CO and CO2 (cracking, dehydration, and decarboxylation) in this
study, are also consistent with the products observed by Lietti et al.
[23,26]. To gain further insight into the reaction pathways, and also
to understand the role of hydrogen, a similar series of experiments
at different W/F were performed in a hydrogen atmosphere.
in the yield of these products at longer W/F suggests that these
compounds may be further converted to secondary products. The
changes in product selectivities as a function of W/F (Fig. 2(b))
further highlight the decrease in lights at higher W/F with a cor-
responding increase in the C6–C9+ coupling products.
The yield of 3-pentanone in H2 is comparable to that obtained
in He. The yield of 2-methylpentanal, that is formed by the hydro-
genation of 2-methyl-2-pentenal, is increased while there is a
significant decrease in the yields of 2-methyl-2-pentenal and the
C6–C9+ ketonization products.
3.4. Proposed reaction network
3.3. Product distribution as a function of W/F at 400 ◦C in
hydrogen
In order to better understand the possible secondary reactions
that modify the product distribution, some of the possible interme-
diate products and co-products were fed over Ce0.5Zr0.5O2 in He or
H2 flow under similar reaction conditions as those used with the
propanal feed. These results are summarized in Table 1.
Hydrogen creates a reduced surface of the oxide, where the coor-
dination environment of the surface cation is altered to give a broad
distribution of cation oxidation states. Hydrogen is adsorbed on
the surfaces of ceria–zirconia as chemisorbed H atoms. These H
atoms interact with surface oxygen to form hydroxyl groups on
the surface. These surface hydroxyl groups subsequently interact
with hydrogen and desorb as H2O, creating an oxygen vacancy and
the reduction of Ce4+ species. FT-IR spectroscopic studies [45] have
shown that there is an increase in the intensity of the OH band up to
the reduction temperature, above which there is a decrease due to
the formation of water and oxygen vacancies on the surface. In the
same study, results from magnetic balance experiments showed
that the reduction percentage of ceria was much higher at a lower
temperature for the mixed oxides, compared to the pure oxide [45],
implying that the mixed oxides are more easily reduced compared
to the pure oxide. The removal of oxygen in this process creates
Lewis acidity [45] that might be helpful for ketonization. In addi-
tion, hydrogen can also promote hydrocracking or hydrogenolysis
reactions. Also, if ketones and aldehydes are hydrogenated, dehy-
dration/hydrogenation is an effective deoxygenation path (in this
case undesirable) that results in generation of light hydrocarbons.
Fig. 2 shows the product distribution obtained at different W/F
over the Ce0.5Zr0.5O2 catalyst under hydrogen flow. As can be
seen from Fig. 2(a), the yield of light hydrocarbons is increased
by a significant amount in the presence of hydrogen. The lights
reach a maximum between W/F of 0.2 and 0.4 and then decrease
gradually with increasing W/F. As mentioned above, it is easy to
see that hydrogen can favor formation of light hydrocarbons and
oxygenates, through cracking and dehydration but the decrease
3.4.1. Feeding oxygenates
Experiments using 2-methyl-2-pentenal as the feed, with either
H2 or He as the carrier, showed that the resulting major products
from this dimer were 2-methylpentanal and light hydrocarbons,
with very low yields of the heavier ketones (Table 1). While the
overall conversion of this C6 aldehyde was much higher under H2
than under He for the conditions tested, the products obtained with
H2 and He were comparable to those obtained with propanal. The
higher yields of lights and 2-methylpentanal obtained with H2 are
also comparable to those with propanal as the feed. This result sug-
gests that the majority of the lighter compounds are formed from
2-methyl-2-pentenal and also confirms the reason for the decrease
in the yield and selectivity of 2-methyl-2-pentenal as a function of
W/F in H2 and in He, as shown in Figs. 1 and 2.
An experiment with 3-pentanone as the feed in He showed
that the main reaction products were 2-butanone, 3-methyl-
2-butanone, 2-methyl-3-pentanone, 3-heptanone, 4-methyl-3-
heptanone and other heavier ketones. The conversion of this C5
ketone was much lower than either propanal or 2-methyl-2-
pentenal under the tested conditions, as shown in Table 1. The
ketones in the products from the reaction of 3-pentanone are com-
parable to those obtained from propanal. This exhibits the ability of
3-pentanone to participate in additional condensation reactions to
give the observed products and also explains the gradual decrease
observed in its selectivity as a function of W/F in He (Fig. 1(b)).

84
A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
Table 1
Product distribution from the reaction of oxygenates over Ce0.5Zr0.5O2 on a COx and H2O free basis, TOS 15 min.
Feed
Carrier gas
He
Temperature, ◦
400
C
W/F, h
0.37
Conversion (%)
42.6
Products
Selectivities (% wt)
2-Methyl-2-pentenal
Lightsa
2-Methylpentanal
Heavies
53.2
24.2
22.5
2-Methyl-2-pentenal
3-Pentanone
H2
400
400
0.37
0.40
96.3
13.0
Lightsa
2-Methylpentanal
Heavies
69.9
20.0
10.0
He
Lightsa
17.5
21.5
9.1
2-Methyl-3-pentanone
2-Butanone
Other coupling ketonesb
51.9
a
Hydrocarbons and oxygenates.
3-Heptanone, 4-methyl-3-heptanone, 4-methyl-3-heptenone and other higher ketones.
b
The results from the experiments with 2-methyl-2-pentenal
oxygenated compounds on several mixed oxide catalysts by tak-
ing up oxygen by an oxidative dehydrogenation (ODH) reaction
[24]. However, no appreciable conversion was observed under the
conditions tested, further confirming that the light oxygenates (2-
methylpropanal, 2-butanone, and 3-methyl-2-butanone) are the
reacting species.
Based on these results, the reaction network shown in Fig. 3
is proposed to account for the formation of the products observed
experimentally. The network contemplates the contribution of two
major reactions, aldol condensation and ketonization, which in
turn involve various condensation steps. In addition, there are also
several side reactions that can take place in parallel. First, it is
and 3-pentanone as the feeds indicate that the same types of chem-
ical reactions and catalytic functions are operative with each of
the feeds over this catalyst. The presence of significant amounts of
hydrocarbons from the reactions of 3-pentanone and 2-methyl-2-
pentenal shows that hydrogenation to the corresponding alcohols
and subsequent dehydration to give the light hydrocarbons is
occurring.
Experiments with ethylene and propylene fed to this catalyst
were conducted to test the reactivity of the light olefins towards
incorporation into oxygenates. Previous studies have shown that
light gases like ethane, ethylene, and propylene can convert to
Fig. 3. Proposed reaction pathway of propanal conversion over Ce0.5Zr0.5O2.

A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
85
known that aldol condensation can occur on both acid and basic
sites [25]. In the mixed oxides, the exposed cations are Lewis acid
sites while the oxygen anions can act as either Lewis or Bron-
sted base sites. Aldol condensation of aldehydes does not require
multiple coordination of surface cations, but involves the reaction
of an adsorbed conjugate base anion with an adsorbed molecule
(ion-molecule reaction) to form a ␤-hydroxyaldehyde that subse-
quently dehydrates to give the higher aldehydes or ketones [12].
The aldol condensation of propanal via the hydrated dimer (3-
hydroxy-2-methylpentanal), produces 2-methyl-2-pentenal that
is subsequently hydrogenated to 2-methyl-pentanal and may
also crack to give some lighter compounds. The aldol condensa-
tion can proceed further to give the cyclic trimers of propanal
that include mesitylene, 2,4,6-trimethylphenol and 2-ethyl-3,5-
dimethylcyclopent-2-en-1-one. However, the trimers are only
minor products and are included here for completeness.
Fig. 4. Evolution of products from pre-adsorbed propanal over Ce0.5Zr0.5O2 as a
In addition to producing the dimers and trimers of propanal,
aldol condensation can also give rise to several other products
observed here. Studies over a ZnCrOx catalyst [23,26] with C3–C4
oxygenates as the feed have reported a wide range of aldehydes
and ketones as the products. It was also shown in that study that
some of the short aldehydes and ketones produced could react fur-
ther to give longer chain products. They have also classified the
reactivities of these short carbonyl species based on their molec-
ular structure and have considered them either as electrophilic or
nucleophilic reactants. Hence, as also included in Fig. 3, the source
of the formation of some of the observed products can be explained
as follows: 3-heptanone from propanal + 2-butanone, 4-methyl-
3-heptanone from propanal + 3-pentanone, 3-methyl-2-butanone
from formaldehyde + 2-butanone, 2-methyl-3-pentanone from 3-
pentanone + formaldehyde. It is to be noted that in the above
reactions, ketones act as the nucleophilic species. The formation
of 2-butanone and formaldehyde will be discussed later.
It is proposed that the ketonization pathway proceeds with oxi-
dation of propanal on the surface to give a propionate intermediate,
which then couples to give the symmetric ketone (3-pentanone).
Oxidation reactions are common on the surfaces of oxides [25,27].
The oxide anions, which are Lewis or Bronsted bases, can trans-
form the adsorbed carbonyl groups to the corresponding surface
carboxylates by nucleophilic attack at the carbonyl carbon of the
substrate by surface oxygen according to the following:
function of temperature, using He as carrier gas.
shown that the pathway to 4-heptanone involves the oxidation of
the alcohol to the aldehyde and subsequently to the acid, which
in turn forms surface carboxylates that couple to give the ketone.
These authors also reported that the oxygen required for the oxi-
dation steps comes from ceria. In addition, Kobune et al. [28] have
reported the formation of 3-pentanone and the higher coupled
ketones on CeO2 and CeO2–Fe2O3 catalysts and mentioned vari-
ous stepwise condensation reactions to form the higher ketones.
Hence, the presence of some ketones observed as the products in
this study can be explained by the following ketonization reactions:
3-pentanone from propanal + propanal, 4-methyl-3-heptanone
from propanal + 2-methylpentanal, 3-methyl-3-pentanone from
propanal + 2-methylpropanal and so on.
The presence of light oxygenates like 2-butanone, 2-
methylpropanal and 3-methyl-2-butanone in the product stream
from the conversion of propanal shows that, in addition to aldol
condensation and ketonization, there can be other side reactions
taking place to give these products. Lietti et al. [23] have suggested
that a C1 oxygenated intermediate, possibly formaldehyde, could
be involved in the formation of these products, but that product has
not been detected. The origin of the formaldehyde species could
be from the hydrogenation of CO2, produced from ketonization
reactions by surface hydrides.
The temperature-programmed desorption (TPD) study of
propanal adsorbed over ceria–zirconia is shown in Fig. 4. Three
main peaks, between temperature T = 400–500 ◦C are evident and
are associated with methane, CO2 and CO. The evolution of methane
and CO2 at the same temperatures indicates the possibility of
surface formate species being involved in the reaction. The forma-
tion of surface formate species over ceria-based catalysts has been
reported by Barteau and co-workers [11]. Hence, the formation of
2-butanone probably occurs via an ␣-addition reaction between
propanal and formaldehyde, while 2-methylpropanal is formed
via a cross-aldol condensation between propanal and formalde-
hyde and 3-methyl-2-butanone is formed through a reverse
␣-addition between 2-methyl-propanal and formaldehyde, as indi-
cated in Fig. 3. The desorption of water can be observed between
T = 150–500 ◦C, as a by-product of aldol condensation, ketonization
and dehydration reactions.
R1–CHO + R2–CHO + 2O2−(s)
→ R1–COO−(s) + R2–COO−(s) + 2Hads + 2e−
Further reaction of these adsorbed carboxylates to form the cou-
pled ketones occurs by:
R1–COO−(s) + R2–COO−(s) + 2Hads→ R1(C O)R2+CO2+H2O+2e−
The above reaction is highly dependent on the availability of
potential reaction partners, which in turn is dependent on the coor-
dinationenvironment ofthesurfacecation andthe redoxproperties
of the oxide. Ceria has the highest surface oxygen mobility com-
pared to other single oxides and the incorporation of zirconia into
the lattice of ceria enhances this surface oxygen mobility and thus
the reducibility of the mixed oxide [16,18,19].
Therefore, with the improvement in the oxygen storage capac-
ity of the mixed oxide, the tendency to form surface carboxylates to
give coupled ketones on Ce0.5Zr0.5O2 is greater than that expected
for a pure oxide. The oxygen vacancy created after decarboxyla-
tive ketonization is replenished by the transport of oxygen from
the bulk, according to the Mars Van Krevelen mechanism. This
property of the mixed oxide is also believed to play an important
role in catalyzing the reactions to the higher ketones. Plint et al.
[27] have proposed such a carboxylate mechanism for forming 4-
heptanone from 1-butanol over a CeO2/MgO catalyst. They have
3.4.2. Effect of addition of acid and water
The role of acids in the overall reaction pathway was examined
using propionic acid as a model feed. Self-condensation of acids on
oxide catalysts gives symmetric ketones, while cross-condensation
between two different acids can give two symmetric ketones (one
from each acid) as well as an asymmetric ketone [29,30]. A number
of studies have shown that a wide range of oxides, including CeO2,

86
A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
Table 2
Effect of incorporating water or propionic acid in the feed on product distribution over Ce0.5Zr0.5O2 catalyst, He gas carrier, 400 ◦C, 1 atm, TOS 15 min.
Feed
Propanal
Propanal + propionic acid
Propanal + water
W/F, h
0.35
–
0.38
5–1
0.35
4–1
Feed: additive ratio (by wt)
Aldol products/3-pentanone ratio
Aldol products/coupling ketones ratio
1.4
0.5
0.5
0.3
0.5
0.2
Product yields (%wt)
Lights
5.6
4.7
4.3
Aldol products
2-Methylpentanal
2-Methyl-2-pentenal
1,3,5-Trimethyl benzene
2,4,6-Trimethyl phenol
2-Ethyl-3,5-dimethyl-2-cyclopent-2-en-1-one
2.3
9.4
0.1
0.6
0.3
2.2
6.8
0.1
0.8
0.3
2.3
5.2
0.1
0.8
0.2
Coupling ketones
3-Pentanone
3-Heptanone
4-Methyl, 3-heptanone
2-Methyl, 3-pentanone
4,4,6-Trimethyl-2-cyclohexen-1-one
Other coupling products
9.0
1.8
1.6
1.0
0.9
9.1
19.0
2.4
1.9
1.4
1.2
15.5
3.3
2.1
1.6
1.1
11.4
13.2
are active for ketonization of acids [31–33]. Some studies have also
reported that solid solutions of ceria-based oxides were the most
active for the ketonization of acids [7,30].
ketones, shown in Table 2 decrease by a significant amount as com-
pared to the feed with pure propanal. This implies that there is
competitive adsorption of the aldehyde and acid on the surface, on
which the acids are preferentially adsorbed. As a result, the num-
ber of sites available for the adsorption and reaction of propanal are
reduced, producing its lower conversion and fewer aldol products.
In agreement with this description, an additional experiment with
a mixed feed of 1:1 mass ratio of acid and aldehyde gave an even
higher proportion of ketonization products compared to aldol prod-
ucts, i.e., a ratio of about 16, while the overall propanal conversion
was further reduced.
From previous studies in the literature, it has been shown that
acids undergo self and cross-condensation reactions to give sym-
metric and asymmetric ketones at relatively low temperatures [29].
Thus, it is expected that with the addition of acids there will be a
significant increase in the yield of 3-pentanone (symmetric ketone
from propionic acid). However, the results show that acids can also
undergo other parallel reactions leading to an increase in the yields
of the coupled ketones (Table 2). Nagashima et al. [29] have studied
the reactivity of propionic acid with various branched and lin-
ear carboxylic acids over CeO2 based mixed oxides. They reported
that with propionic acid and another carboxylic acid, three ketones
were obtained; 3-pentanone from propionic acid, the symmetric
ketone from the other acid, and the corresponding asymmetric
(cross-condensation) ketone. This is in agreement with our results,
where with the addition of propionic acid, we see an increase in
the yield of the symmetric ketone as well as an increase in the
yields of the other coupling products. These coupling products are
mainly higher ketones and are grouped together in the table for
simplicity.
Fig. 5 shows the conversion of propanal as a function of TOS,
comparing an experiment using pure propanal as feed to another
one using a mixed feed of propanal and propionic acid (5:1 mass
ratio). No acids were detected in the products in the co-feed exper-
iments, showing that the reaction of the acid on the surface of the
oxide to form ketones is very fast. Fig. 5 also shows that, with the
co-feed of propionic acid, the conversion of propanal was reduced
significantly. The deactivation of the catalyst was more severe for
the feed with acid incorporated than for the pure propanal feed.
Table 2 shows the yields of the products obtained in both experi-
ments at the shortest TOS. Here, the aldol products are the dimers
and trimers from the self-condensation of propanal, while the cou-
pled ketones are all the other observed ketones. The results show
a significant increase in the yield of 3-pentanone with the addition
of propionic acid and also an increase in the yields of the coupled
ketones. However, a decrease in the yields of 2-methyl-2-pentenal
(aldol dimer) and lights are observed. Also, the ratio of aldol prod-
ucts to 3-pentanone and the ratio of aldol products to the coupled
Hence, the observed increase in the yields of ketones upon
acid incorporation can be attributed to an increase in the amount
of carboxylate type species on the surface and also results in
an increase in the amount of CO2 produced from the increased
ketonization.
With more CO2 produced, there is a possibility of forming
higher amounts of formaldehyde and subsequently increases in the
amounts of 2-butanone and 2-methylpropanal. These species can
then undergo aldol condensation and ketonization reactions, as dis-
cussed in the previous section, and hence contribute to an increase
in the yields of the higher ketones. However, with an increase in
the carboxylate type species on the surface, it is possible that these
can promote a variety of cross-condensation reactions, in addition
Fig. 5. Effect of addition of acid on propanal conversion over Ce0.5Zr0.5O2. Reaction
conditions: 400 ◦C, 1 atm, He.

A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
87
condensation products. The ratios of aldol products to 3-pentanone
and the coupling ketones also show a decrease with the feed with
water.
Water is adsorbed on the surface of oxides as H and OH groups
(dissociative chemisorption). These surface hydroxyl groups are
moieties that can initiate nucleophilic attack by oxygen at the
carbonyl carbon of the aldehydes to form the corresponding
acids [24,35]. In the presence of water with propanal over the
ceria–zirconia catalyst, the surface hydroxyl groups can increase
the transformation of the adsorbed propanal to the surface propi-
onate species, as depicted in Fig. 7. The surface propionate can then
couple with another similar species and desorb as the symmetric
ketone or react further with other surface species via the pathways
shown in Fig. 3 to give the observed increase in the yields of the
products in the presence of water. However, it should be noted that
with increasing partial pressure of water, it is very likely that the
reaction would eventually be inhibited by blocking of the active
sites by adsorbed water molecules. Therefore, the effect of water
on the ketonization activity is a balance between the availability
of O2− anions as well as the extent of hydroxylation of the surface,
since an oxygen anion is required for the abstraction of an H atom
to form the propionate species. Yokoyama et al. [35] have reported
that, with benzaldehyde and water vapor in the feed over ZrO2 and
Cr2O3/ZrO2 catalysts, benzaldehyde forms a benzoate species on
the surface with the lattice oxygen. A surface hydroxyl group then
attacks the carboxyl group of the benzoate species to produce ben-
zoic acid and hydrogen as the observed products. That study further
supports the results observed here.
Fig. 6. Effect of water on propanal conversion over Ce0.5Zr0.5O2. Reaction conditions:
400 ◦C, 1 atm, W/F 0.35 h, He, Fpropanal 0.2 ml/h, Fwater 0.04 ml/h.
to self-coupling that forms the symmetric ketone. This provides an
explanation for the observed increase in yield of coupling products,
as well as their complexity.
Water is
a byproduct from both aldol condensation and
In summary, these results have shown the integral role played
by acids and water in altering the reaction pathway; the addi-
tion of acids or water suppresses aldol condensation and promotes
ketonization. While addition of acid reduces the conversion of
propanal, addition of water increases the conversion and enhances
the catalyst stability over time.
ketonization reactions and is believed to play an important role in
promoting ketonization reactions on oxides. Studies of oxidation
of hydrocarbons on metal oxide catalysts [24] and benzaldehyde
reactions on oxide catalysts [35] have shown that water enhances
the oxidation of the aldehydes to the corresponding acids. This is
because, in the presence of water molecules, the concentration
of the surface hydroxyl species is increased and these hydroxyl
species facilitate the attack at the carbonyl carbon of the substrate
by the lone pair of electrons on the oxygen atom to form the corre-
sponding carboxylates on the surface that subsequently desorb as
acids.
Fig. 6 shows the conversion of propanal as a function of TOS
for feeds with and without water. It can be seen that the conver-
sion and stability of the catalyst are improved with the addition
of water. The improvement in stability may be because water
acts as a source of oxygen that helps to desorb products from
the surface and also promotes the oxidation of coke, thus creat-
ing more active sites for the reaction to proceed. Table 2 shows
the yields of the products obtained for this experiment, together
with the results for the experiments with pure propanal and the
propionic acid co-feed experiment. A similar effect, as with the
addition of acid, is observed here with water. There is a signif-
icant increase in the yield of 3-pentanone and also an increase
in yields of the C6–C9+ coupling products with a corresponding
decrease in the yields of 2-methyl-2-pentenal and lights, while
there is not much change in the small yields of the higher aldol
3.4.3. XPS analysis of fresh and spent catalysts
The reaction experiments performed thus far have suggested
that the ketonization of propanal over the ceria–zirconia mixed
oxides occurs through the oxidation of propanal on the surface to
form surface carboxylates and further reactions of these carboxy-
late type species give 3-pentanone and higher ketones. The key
step to these reactions, as discussed in the previous sections, is the
formation of the carboxylate intermediate on the surface and this
suggests that the oxygen from the surface of the mixed oxide partic-
ipates in this reaction. In doing so, the cerium would shift between
the Ce4+ and Ce3+ states to create oxygen vacancies. Therefore, it
is of interest to determine the nature of the surface and the oxi-
dation states of the elements before and after reaction. Therefore,
a series of experiments were performed with the catalyst samples
being treated under different conditions. The catalyst samples ana-
lyzed were: (i) fresh, (ii) reduced at 400 ◦C, (iii) after reaction of
propanal in H2, (iv) after reaction of propanal in He, (v) after reac-
tion of propanal with co-feed of propionic acid, (vi) after reaction
of propanal with co-feed of water.
Fig. 7. Schematic of the interaction of adsorbed water with propanal.

88
A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
Fig. 8. (a) Ce (3d) and (b) Zr (3d) XPS spectra for Ce0.5Zr0.5O2 treated under different conditions: (i) fresh catalyst, (ii) catalyst reduced at 400 ◦C, (iii) reaction in H2, (iv)
reaction in He, (v) reaction with co-feed of acid, and (vi) reaction with co-feed of water.
Fig. 8 shows the XPS spectra of Ce (3d) and Zr (3d) regions of
the samples from these experiments. It can be observed that the Ce
(3d) spectrum is composed of several peaks and the complexity of
the spectrum is reported as being due to the hybridization between
the partially occupied 4f levels of ceria and the 2p states of oxygen
[19,46–49]. The spectrum was fitted with up to eight peaks with
Gaussian distributions and the main peaks and satellites were iden-
tified and labeled according to those used by Burroughs et al. [14].
Peaks denoted as u and v correspond to 3d3/2 and 3d5/2 contribu-
tions, respectively, while peaks with primed labels denote satellite
features. The peaks v, vꢀꢀ, and vꢀꢀꢀ (and correspondingly for u) are
attributed to Ce4+, while vꢀ and uꢀ are attributed to Ce3+. The spec-
trum of the Zr (3d) region shown in Fig. 8(b) consists of Zr 3d3/2
and Zr 3d5/2 primary ionization features and was fitted with two
Gaussian curves.
In Fig. 8(a), it can be seen that there is a shift to higher bind-
ing energies in the Ce (3d) spectrum as the treatment conditions
change from fresh catalyst to the one with reaction in He. There also
seems to be a slight shift to lower binding energies for the one with
the co-feed of water. This shift in the positions of the Ce (3d) band
is indicative of a change in the coordination environment of the
Ce atoms and could possibly be due to a combination of electronic
and morphological changes. Therefore, it is conservative to deter-
mine the oxidation states on the surface on the basis of the method
reported in a number of works in the literature [19,46,47]. By this
method, the ratio of the area of uꢀꢀꢀ satellite peak relative to the total
area under the Ce 3d spectrum was calculated, since the uꢀꢀꢀ does
not overlap with the other peaks in the spectrum. This ratio then
corresponds to the % Ce4+ (or Ce3+) in the samples. Table 3 shows
the % uꢀꢀꢀ and the % of Ce3+ estimated from the XPS analysis of the
samples. It can be seen that the fresh ceria–zirconia catalyst had the
lowest value of the % Ce3+ of 7 and the reduction of this sample in
H2 at 400 ◦C increased this value to 10. The increase in the value of %
Ce3+ is expected since reduction in H2 creates oxygen vacancies and
decreases the oxidation state of cerium. The reaction of propanal in
H2 at 400 ◦C increased this value to 29, while for the same reaction
under He, the increase was only to 18. The differences observed
with H2 and He are explained in terms of the reacting environment
under the two carrier gases. The surface is more deficient in oxy-
gen with H2 as the carrier and with the reactions of propanal, both
consume oxygen (H2 creates oxygen vacancies, while reacting sur-
face species require coordination to oxygen) and therefore create
lower oxidation state Ce ions. However, with He as the carrier, the
oxygen is removed only by the reacting species and the amount of
Ce4+ reduced is lower.
With the co-feeds of water and propionic acid, it is expected
that the reduction of the oxide will be less compared to the pure
propanal feed, as the surface will be enriched in oxygen by the pres-
ence of oxygen containing molecules (propionic acid and water) in
the feed, in addition to propanal. From Table 3, it is clear that the
reaction with co-feed of water exhibited this effect, as the value of
% Ce3+ is much lower (14) compared to the other conditions and is
closer to the value of the reduced sample. This is consistent with
the results reported in Section 3.4.2 for co-feed of water and sup-
ports the proposed observation that water promotes the oxidation
of propanal. With the co-feed of acid, the increase in the value of %
Ce3+ (23) could be due to further reactions initiated in the presence
of acids on the surface, leading to a more reduced surface of the
oxide. This is also in agreement with the results from Section 3.4.2,
where with the addition of acid in the feed, there was an increase in
the yields of the coupling products. In summary, these results have
shown that the state of reduction of the mixed oxide under differ-
ent reaction conditions support the results observed in the steady
state runs.
3.4.4. Effect of changing the composition of the mixed oxides
Table 3
Percent of uꢀꢀꢀ and Ce3+ from XPS analysis.
The changes in acid–base properties of mixed oxide cata-
lysts with the change in the composition have been reported
in previous studies [7,32–34,37,38]. The incorporation of ZrO2
into the fluorite structure of CeO2 can significantly change the
acid–base properties of the mixed oxide in addition to improv-
ing the oxygen mobility through the lattice [13,15–19,40,41]. This
property of the mixed oxide catalyst has been studied here with
the condensation reactions of propanal and the results obtained
are discussed based on the structural properties of the different
catalysts.
uꢀꢀꢀ in Ce (3d)
(area%)
Atomic Ce3+
(%)
Fresh CZ 50
12.26
11.83
9.29
10.84
10.08
11.37
7
10
29
18
23
14
CZ 50 reduced at 400 ◦C
CZ 50 reaction in H2
CZ 50 reaction in He
CZ 50 reaction with co-feed of acid
CZ 50 reaction with co-feed of water
CZ 50—Ce0.50Zr0.50O2.

A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
89
Ce–Zr (50–50) catalyst shows a balance in catalyzing both aldol
condensation and ketonization reactions. In addition, the compari-
son between aldol condensation and ketonization was tested with
a non-reducible oxide, ␥-alumina. The results obtained with ␥-
alumina are comparable with those obtained with ZrO2 and Ce–Zr
(25–75), as shown in Table 4. The major product observed is the
aldol dimer, 2-methyl-2-pentenal, with only small yields to 3-
pentanone.
The results obtained here show that, by varying the ceria
and zirconia content of the mixed oxide catalyst, the selectivity
towards either aldol condensation or ketonization can be increased.
This can be explained in terms of the nature of the active sites
required for the reactions. Aldol condensation requires a basic site
for the extraction of an ␣-hydrogen to form an enolate species
and then subsequent dimerization giving the aldol dimer and
other condensation products. FTIR studies of CO2 adsorbed on
ceria–zirconia mixed oxide catalysts [42] were performed to show
that ceria–zirconia mixed oxides possess weak, medium and strong
basic sites that corresponding to OH−, M–O2− and low coordinat-
ing O2− groups, respectively. In addition, it has also been shown
that pure ZrO2 has only surface OH− and M-O2− groups [42,43].
Therefore, it is probable that the aldol condensation reactions are
catalyzed by the surface OH− and M–O2− groups on the mixed
oxides and that with an increasing amount of zirconia, there is a
greater tendency to form the aldol condensation dimer, 2-methyl-
2-pentenal from propanal, as shown in Table 4. In addition, an
increasing amount of zirconia also leads to an increase in the yields
to aromatic trimers from propanal (Table 4). This is due to an
increase in the density of acid sites with increasing zirconia and
the aromatization reactions being catalyzed by the acid sites [44].
This is also consistent with the results observed with ␥-alumina
where, 2-methyl-2-pentenal was the major observed product, with
significant yields to the aromatic trimers (Table 4).
The formation of 3-pentanone over ceria–zirconia was dis-
cussed in the previous sections and is believed to proceed through
the coupling of surface carboxylates. The extra oxygen necessary
for this step comes from the lattice of the mixed oxide. How-
ever, 3-pentanone is also observed in the products over zirconia
and ␥-alumina, which are not known to be reducible like ceria
or ceria–zirconia. In addition, CO2 and H2 were observed as the
gas phase product over zirconia and ␥-alumina, while only minor
amounts of CO were detected. This leads to the question of the
source of the extra oxygen that is required to form the carboxylates
on the surface, leading to decarboxylative ketonization. Previous
studies in literature have reported the possibility of the Canniz-
zaro reaction [51], decomposition from esters [50] and coupling
Fig. 9. Effect of changing the composition on propanal coversion over (a) CeO2, (b)
Ce0.75Zr0.25O2, (c) Ce0.5Zr0.5O2 and (d) Ce0.25Zr0.75O2 (e) ZrO2. Reaction conditions:
400 ◦C, 1 atm, W/F 0.35 h, He.
The catalysts tested were ZrO2, Ce0.25Zr0.75O2, Ce0.5Zr0.5O2 and
Ce0.75Zr0.25O2 and CeO2. The mixed oxides will be referred to as
Ce–Zr (25–75), Ce–Zr (50–50) and Ce–Zr (75–25), respectively.
Fig. 9 shows the conversion as a function of TOS over the catalysts
under helium flow. It can be observed that, with the same space
time (W/F) of 0.35 hr, pure zirconia exhibited the highest activity
for the conversion of propanal, followed by Ce–Zr (25–75) > Ce–Zr
(50–50) > Ce–Zr (75–25) > pure ceria. The higher activity observed
with catalysts with higher zirconia content could be due to the
nature of the sites present in the catalyst that favor either aldol
condensation or ketonization reactions. This will be discussed in
detail below. The deactivation profile looks similar for all the cat-
alysts over a 4 h reaction period. Table 4 compares the product
yields for the reaction of propanal over the five catalysts. The space
time (W/F) was adjusted so that a similar level of conversion was
achieved for all the catalysts for comparison. With a change in the
composition of the oxides, a significant change in the product dis-
tribution is observed. In Table 4, from left to right, the catalysts
have increasing zirconia content (reduced ceria content) and it
can be seen that with an increasing amount of zirconia, there is
a significant increase in the yields to 2-methyl-2-pentenal (aldol
dimer) and other aldol condensation trimers with a corresponding
decrease in the yields to 3-pentanone. Similarly, with an increasing
ceria content (right to left in Table 4), the 3-pentanone formation
is promoted while aldol condensation reactions are reduced. The
Table 4
Product distribution from conversion of propanal over CexZr1−xO2 and ␥-Al2O3at 400 ◦C, 1 atm, TOS 15 min.
Feed
Propanal
CeO2
Catalyst
Ce0.75Zr0.25O2
Ce0.50Zr0.50O2
Ce0.25Zr0.75O2
ZrO2
0.038
␥-Al2O3
W/F, h
Conversion (%)
1.02
32.7
0.42
38.4
0.36
41.5
0.22
43.8
0.36
49.0
41.8
Temperature, ◦C
400
Ratio 3-pentanone/dimer
3.05
2.92
0.95
0.20
0.08
0.23
Product yields (%wt)
Lights
3-Pentanone
2-Methylpentanal
7.9
8.4
1.2
6.9
12.3
2.5
5.6
9.0
2.3
4.7
4.7
3.4
3.6
1.7
2.1
8.7
4.6
3.8
2-Methyl-2-pentenal
2-Ethyl-3,5-dimethyl-2-cyclopent-2-en-1-one
1,3,5-Trimethyl benzene
2,4,6-Trimethyl phenol
Other condensation products
2.7
0.0
0.0
0.9
4.2
0.0
0.0
0.5
9.4
0.3
0.1
0.6
23.3
1.8
0.4
0.5
4.9
22.5
3.3
1.0
0.5
7.0
20.3
4.9
1.3
0.0
5.2
11.6
12.1
14.4

90
A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
of carboxylates [27] as possible ways to form ketones over oxides.
The possibility of forming ketones from the decomposition of an
ester was tested by co-feeding 1-propanol and 1-butanol with
propanal over zirconia and ␥-alumina in separate experiments.
The alcohol and the aldehyde are expected to form an ester on
the surface through a hemiacetal intermediate with the ester sub-
sequently decomposing to give the ketone [50]. However, the
results observed here showed no significant changes in the yields
of 3-pentanone in case of 1-propanol and gave only minor yields
to the expected ketone, 3-hexanone with 1-butanol. This shows
that the contribution of this pathway to form the ketone is mini-
mal for the conditions of this study. If the Cannizzaro reaction is
expected to proceed to form the ketone, then an enhanced forma-
tion of 1-propanol should be expected, which is also not the case
here. This observation is consistent with the spectroscopic studies
of benzoyl compounds on oxides by Ponec and co-workers [52],
where the bands related to the expected Cannizzaro reaction prod-
ucts of phenylmethylketone were not observed over ␥-alumina.
Instead, bands assignable to methoxy and benzoate species were
observed. This implies that the bond between the methyl and
carbonyl group in phenylmethylketone is broken and the surface
oxygen participates in the transformation of these groups to form
the observed methoxy and benzoate species. In the same study,
with benzaldehyde as the feed, they have also mentioned that
the transformation to a surface benzoate occurs with the abstrac-
tion of hydrogen as the first step. This is then followed by the
nucleophilic attack at the carbonyl group by surface oxygen to
form the benzoate species. This leads to the possibility that a
similar transformation is also occurring on the surfaces of zirco-
nia and ␥-alumina in this study to form 3-pentanone and thus
also explaining the formation of CO2 and H2 that is observed in
the gas phase. However, the non-reducible nature of these oxides
explains the low yields of the ketones as compared to ceria or
ceria–zirconia.
increased Ce favored ketonization. This occurred by shifting the
balance of the acid–base properties of the active sites.
Acknowledgements
Support from the National Science Foundation (EPSCoR
0814361), US Department of Energy (DE-FG36GO88064), Okla-
homa Secretary of Energy and the Oklahoma Bioenergy Center are
greatly appreciated. The authors would like to thank Dr. Rolf Jentoft
for the help with the X-ray diffraction measurements.
References
[1] P. Mckendry, Bioresour. Technol. 83 (2002) 47–54.
[2] C.A. Mullen, A.A. Boateng, Energy Fuels 22 (2008) 2104–2109.
[3] A. Corma, G.W. Huber, L. Sauvanaud, P. O’ Connor, J. Catal. 257 (2008) 163–171.
[4] M. Glinski, J. Kilenski, Appl. Catal. A 190 (2000) 87–91.
[5] K. Okumura, Y. Iwasawa, J. Catal. 164 (1996) 440–448.
[6] K.S. Kim, M.A. Barteau, J. Catal. 125 (1990) 353–375.
[7] Y. Kamimura, Y. Kamimura, S. Sato, R. Takahashi, T. Sodesawa, T. Akashi, Appl.
Catal. A 252 (2003) 399–411.
[8] T.S. Hendren, K.M. Dooley, Catal. Today 85 (2003) 333–351.
[9] K.M. Dooley, A.K. Bhat, C.P. Plaisance, A.D. Roy, Appl. Catal. A 320 (2007)
122–133.
[10] S.D. Randery, J.S. Warren, K.M. Dooley, Appl. Catal. A 226 (2002) 265–280.
[11] H. Idriss, C. Diagne, J.P. Hindermann, A. Kiennemann, M. Barteau, J. Catal. 155
(1995) 219–237.
[12] A. Trovarelli, C. de Leitenmburg, G. Dolcetti, Chemtech 27 (1997) 32–37.
[13] C.A. Gaertner, J.C. Serrano-Ruiz, D.J. Braden, J.A. Dumesic, J. Catal. 266 (2009)
71–78.
[14] P. Burroughs, A. Hammett, A.F. Orchard, G. Thornton, J. Chem. Soc., Dalton Trans.
17 (1976) 1686–1698.
[15] A. Trovarelli, C. de Leitenburg, M. Boaro, G. Dolcetti, Catal. Today 50 (1999)
353–367.
[16] J.I. Gutiérrez-Ortiz, B. de Rivas, R. López-Fonseca, J.R. González-Velasco, J.
Therm. Anal. Calorim. 80 (2005) 225–228.
[17] G. Ranga Rao, P. Fornasiero, R. Di Monte, J. Kaspar, G. Vlaic, G. Balducci,
S. Meriani, G. Gubitosa, A. Cremona, M. Graziani, J. Catal. 162 (1996)
1–9.
[18] C.E. Hori, H. Permana, K.Y. Simon Ng, A. Brenner, K. More, K.M. Rahmoeller, D.
Belton, Appl. Catal. B 16 (1998) 105–117.
[19] F.B. Noronha, E.C. Fendley, R.R. Soares, W.E. Alvarez, D.E. Resasco, Chem. Eng. J.
82 (2001) 21–31.
4. Conclusions
[20] Jeong-Rang Kim, Wan-Jae Myeong, Son-Ki Ihm, Appl. Catal.
57–63.
[21] Y. Kamimura, S. Sato, R. Takahashi, T. Sodesawa, T. Akashi, Appl. Catal. A 253
(2003) 399–411.
B 71 (2007)
The following important conclusions can be drawn from the
current study:
[22] J.B. Claridge, M.L.H. Green, S.C. Tsang, A.P.E. York, J. Chem. Soc., Faraday Trans.
89 (7) (1993) 1089–1094.
[23] L. Lietti, E. Tronconi, P. Forzatti, J. Catal. 135 (1992) 400–419.
[24] M.M. Bettahar, G. Costentin, L. Savary, J.C. Lavalley, Appl. Catal. A 145 (1996)
1–48.
[25] Mark.A. Barteau, Chem. Rev. 96 (4) (1996) 1413–1430.
[26] L. Lietti, E. Tronconi, J. Mol. Catal. A: Chem. 94 (1994) 335–346.
[27] N. Plint, D. Ghavalas, T. Vally, V.D. Sokolovski, N.J. Coville, Catal. Today 49 (1999)
71–77.
[28] M. Kobune, S. Sato, R. Takahashi, J. Mol. Catal. A: Chem. 279 (2008) 10–19.
[29] O. Nagashima, S. Sato, R. Takahashi, T. Sodesawa, J. Mol. Catal. A: Chem. 227
(2005) 231–239.
[30] M. Renz, Eur. J. Org. Chem. 6 (2005) 979–988.
[31] J.C. Kuriacose, R. Swaminathan, J. Catal. 14 (1969) 348–354.
[32] R. Pestman, A. van Duijne, J.A.Z. Pieterse, V. Ponec, J. Mol. Catal. A: Chem. 103
(1995) 175–180.
[33] M. Glinski, J. Kijenski, A. Jakubowski, Appl. Catal. A 128 (1995) 209–217.
[34] T. Akashi, S. Sato, R. Takahashi, T. Sodesawa, K. Inui, Catal. Commun. 4 (2003)
411–416.
[35] T. Yokoyama, N. Fujita, T. Maki, Appl. Catal. A 125 (1995) 159–167.
[36] S. Sato, R. Takahashi, T. Sodesawa, K. Matsumoto, Y. Kamimura, J. Catal. 184
(1999) 180–188.
[37] V.K. Diez, C.R. Apesteguia, J.I. Di Cosimo, J. Catal. 215 (2003) 220–233.
[38] J.I. Di Cosimo, V.K. Diez, M. Xu, E. Iglesia, C.R. Apesteguia, J. Catal. 178 (1998)
499–510.
(a) Propanal is converted to higher carbon chain oxygenates on the
ceria–zirconia catalyst by two pathways, aldol condensation
and ketonization, which is in agreement with previous studies
for other oxygenates and other oxide catalysts.
(b) The important intermediate for ketonization is a surface car-
boxylate formed by oxidation of the aldehyde by the oxidizing
solid surface.
(c) The presence of acids in the feed inhibits the aldol condensation
pathway by competitive adsorption of the acid that reduces the
aldehyde adsorption and conversion.
(d) Water promotes ketonization and inhibits aldol condensation
by increasing the concentration of surface hydroxyl groups that
are active sites and this increases the activity by enhancing the
formation of the surface carboxylates by the aldehyde.
(e) Catalyst stability is also increased by the presence of water.
(f) Hydrogen enhances cracking and produces light oxygenates
and light hydrocarbons. The light oxygenates may be reincor-
porated to give secondary products, but the light hydrocarbons
do not react further.
(g) Analysis of the fresh and spent catalysts by XPS showed varying
degrees of reduction of the oxides under different treatment
conditions that are consistent with oxide oxygen participating
in the reactions.
[39] A. Trovarelli, Catalysis by Ceria and Related Materials, vol. 2, Imperial College
Press, London, 2001.
[40] M.G. Cutrufllo, I. Ferino, R. Monaci, E. Rombi, V. Solinas, Top. Catal. 19 (2002)
225–240.
[41] E.I. Gurbuz, E.L. Kunkes, J.A. Dumesic, Appl. Catal. B 94 (2010) 134–141.
[42] M. Daturi, C. Binet, J. Lavalley, A. Galtayries, R. Sporken, Phys. Chem. Chem.
Phys. 1 (1999) 5717–5724.
[43] D. Bianchi, T. Chafik, M. Khalfallah, S.J. Teichner, Appl. Catal. A 112 (1994) 57–
73.
(h) Changing the proportion of the parent oxides showed that
increased Zr favored formation of aldol products while

A. Gangadharan et al. / Applied Catalysis A: General 385 (2010) 80–91
91
[44] G.S. Salvapati, K.V. Ramanamurty, M. Janardanarao, J. Mol. Catal. 54 (1989)
9–30.
[48] S. Damyanova, B. Pawelec, K. Arishtirova, M.V. Martinez Huerta, J.L.G. Fierro,
Appl. Catal. A 337 (2008) 86–96.
[45] M. Daturi, E. Finocchio, C. Binet, J.C. Lavalley, F. Fally, V. Perrichon, J. Phys. Chem.
B 103 (1999) 4884–4891.
[46] J.Z. Shyu, K. Otto, W.L.H. Watkins, G.W. Graham, R.K. Belitz, H.S. Gandhi, J. Catal.
114 (1988) 23–33.
[49] A.E. Nelson, K.H. Schulz, Appl. Surf. Sci. 210 (2003) 206–221.
[50] M. Glinski, W. Szymanski, D. Lomot, Appl. Catal. A 281 (2005) 107–113.
[51] H. Idriss, K.S. Kim, M.A. Barteau, J. Catal. 139 (1993) 119–133.
[52] C.A. Kouststaal, P.A.J.M. Angevaare, V. Ponec, J. Catal. 143 (1993) 573–582.
[47] J.Z. Shyu, W.H. Weber, H.S. Gandhi, J. Phys. Chem. 92 (1988) 4964–4970.