Selective hydrodeoxygenation of bio-oil derived products: Acetic acid to propylene over hybrid CeO2-Cu/zeolite catalysts

Article abstract of DOI:10.1039/c5cy01485a

Conversion of acetic acid, the light oxygenate from biomass pyrolysis, to propylene can be achieved via keto-hydrodeoxygenation (KHDO) over hybrid CeO₂-Cu/zeolite catalysts at >573 K under atmospheric H₂. The catalyst containing CeO₂ and Cu/HY (25 wt% of Cu/HY) was employed to obtain up to 85% conversion of acetic acid with 49% selectivity to propylene. Acetone, propylene and propane are obtained via ketonization-hydrogenation-dehydration over the three-component catalyst, while ethanol, acetaldehyde, ethylene, ethane and ethyl acetate can also be produced from hydrogenation-dehydration over Cu/zeolites alone. The catalyst containing Cu/HY provides higher selectivity to olefin products, as compared to that containing Cu/HZSM-5. The reaction is suppressed by the presence of water. Nevertheless, high catalyst stability (>60 hours on stream) can be obtained. The KHDO can be applicable for the conversion of acetic acid, a biomass derived product, to hydrocarbons using a sequential bed system of the three-component catalyst and HZSM-5 catalyst.

Full text of DOI:10.1039/c5cy01485a

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Selective hydrodeoxygenation of bio-oil derived
products: acetic acid to propylene over hybrid
CeO2–Cu/zeolite catalysts†
Cite this: DOI: 10.1039/c5cy01485a
a
ab
*
Ayut Witsuthammakul and Tawan Sooknoi
Conversion of acetic acid, the light oxygenate from biomass pyrolysis, to propylene can be achieved via
keto-hydrodeoxygenation (KHDO) over hybrid CeO2–Cu/zeolite catalysts at >573 K under atmospheric H2.
The catalyst containing CeO2 and Cu/HY (25 wt% of Cu/HY) was employed to obtain up to 85% conversion
of acetic acid with 49% selectivity to propylene. Acetone, propylene and propane are obtained via keto-
nization–hydrogenation–dehydration over the three-component catalyst, while ethanol, acetaldehyde, eth-
ylene, ethane and ethyl acetate can also be produced from hydrogenation–dehydration over Cu/zeolites
alone. The catalyst containing Cu/HY provides higher selectivity to olefin products, as compared to that
containing Cu/HZSM-5. The reaction is suppressed by the presence of water. Nevertheless, high catalyst
stability (>60 hours on stream) can be obtained. The KHDO can be applicable for the conversion of acetic
acid, a biomass derived product, to hydrocarbons using a sequential bed system of the three-component
catalyst and HZSM-5 catalyst.
Received 5th September 2015,
Accepted 9th October 2015
DOI: 10.1039/c5cy01485a
www.rsc.org/catalysis
support can increase the degree of oxygen removal but inhibit
the further hydrogenation to hydrocarbons.17 Many noble
1. Introduction
Today, the world's highest demand for petrochemical feed-
stock is small olefins, especially propylene and ethylene.1,2 As
fossil reserves are running low, the biomass-derived products
have become promising future sources. The strategy involves
deoxygenation of small oxygenates to olefins with a minimal
hydrogen consumption. Among those oxygenates that were
derived, acetic acid is a potential feedstock as it can be largely
obtained from biomass pyrolysis3,4 and fermentation of agri-
cultural products and wastes.5 However, the controlled deoxy-
genation of acetic acid to olefins is somewhat challenging.
This is due to high oxygen content in the molecule, as com-
pared to other feedstocks. Many studies have focused on par-
tial deoxygenation via ketonization over metal oxide catalysts
(MgO, CdO, MnO2, and Fe2O3) to acetone, CO2 and water.6–11
Alternatively, the dehydration of acetic acid to ethenone,
which leads to rapid deactivation of the catalyst, was also
reported.12,13 Meanwhile, the hydrodeoxygenation (HDO) of
acetic acid has been investigated over Pt–Sn alloy, Cu and Co
catalysts. In this case, ethanol, acetaldehyde and ethyl acetate
are obtained.14–16 The esterification facilitated by an acidic
metals including Pt, Pd, Ni, and Rh were found to be effective
catalysts for liquid and gas phase hydrogenation of acetic acid
to hydrocarbons.4,15,18,19 However, a high H2 pressure (>4
MPa) or a high temperature (typically >698 K) is required to
obtain appreciable activity over these metals.4,20 In addition,
the hydrocarbons obtained are mainly paraffins (ethane and
methane), presumably due to successive hydrogenation,
decarboxylation and hydrogenolysis over those metals.15,18,21
Due to the challenges mentioned above, a novel approach
for obtaining olefins from acetic acid is proposed in this
work via the keto-hydrodeoxygenation process. Previous work
demonstrated that ketone could be hydrodeoxygenated to ole-
fin via a controlled hydrogenation–dehydrogenation pro-
cess.22 At the same time, acetic acid can be selectively keto-
nized over various metal oxides that are relatively inert for
hydrogenation and dehydration.23,24 Hence, it is possible to
incorporate these metal oxides into the hydrodeoxygenation
catalysts for a single stage conversion of acetic acid to olefins
via ketonization–hydrogenation–dehydration. In this work,
CeO2 will be selected as a ketonization catalyst due to high
activity especially at low temperature.9 Cu/HY and Cu/HZSM-
5, showing high hydrodeoxygenation activity from the previ-
ous work,22 will be incorporated for hydrogenation–dehydra-
tion of the ketone formed. Despite the difference in catalytic
parameters for each step, formulation of the three-
component catalyst (CeO2/Cu/zeolites) that works isother-
mally under relatively mild conditions will be optimized.
a Department of Chemistry, Faculty of Science, King Mongkut's Institute of
Technology Ladkrabang, Chalongkrung Road, Bangkok 10520, Thailand.
E-mail: kstawan@kmitl.ac.th; Fax: +66 2 326 4415; Tel: +66 81 929 8288
b Catalytic Chemistry Research Unit, Faculty of Science, King Mongkut's Institute
of Technology Ladkrabang, Chalongkrung Road, Bangkok 10520, Thailand
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c5cy01485a
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Their effects on the reaction and products are also
highlighted, together with the role of water present in the
acetic acid feed.
found in the work by Sagar et al.28 Acidity of all zeolite sam-
ples was quantified by NH3-TPD. 1% NH3/He was pre-
adsorbed at 323 K. TPD was carried out in He at 10 K min−1
from 323 to 973 K.29 The particle size of Cu on a support was
estimated by TEM using a LaB6 emitter (FEI Tecnai G2 20,
200 kV).
2. Experimental procedure
CeO2 (99.9%) was purchased from Sigma-Aldrich®. HY and
HZSM-5 were commercially obtained from Tohso and
Zeochem®, respectively. The metal oxides and zeolite were
calcined in air at 723 K for 5 h before use. 5 wt% Cu/zeolites
were simply prepared by incipient wetness impregnation.
Briefly, a CuIJNO3)2·3H2O precursor (Ajax Fine Chem) was
dissolved in deionized water (~0.01 M) and slowly dropped
onto the zeolites until wet. The sample was dried at 333 K in
an oven for 15 min, and then the loading was repeated until
the desired metal content was reached. The samples were
allowed to dry at 333 K overnight and calcined in air at 723 K
for 5 h. The CeO2 was physically mixed with Cu/zeolite to
obtain a hybrid CeO2–Cu/zeolite catalyst (7–40 wt% Cu/zeo-
lites) and then pelletized to the size of 600–850 μm. The
three-component catalyst is referred to as CeO2–Cu/zeoliteIJX),
where X is the wt% of Cu/zeolite in the mixed catalyst.
The catalytic testing was conducted using a fixed bed
flow reactor (6 mm i.d. Pyrex®) at atmospheric pressure.
The CeO2–Cu/zeolite catalysts were primarily activated at
723 K (2 K min−1) under a stream of air (30 ml min−1) for 5
h. Subsequently, the catalyst was flushed with N2 and
treated in H2 at 723 K for another 2 h. The system was
cooled down to the reaction temperature (573 K), and the
reaction was carried out at atmospheric hydrogen pressure.
The acetic acid was introduced by using a syringe pump at
the rate of 0.5–1.0 g h−1. The products were analyzed by
using an on-line GC-FID. A Hayesep® P (1/8″ × 8′) was used
as a separating column.
3. Results and discussion
3.1 Catalyst characterization
The elemental composition of the catalysts was deter-
mined by using an X-ray fluorescence spectrometer (XRF; Sie-
mens). The specific surface area (BET) of the catalysts was
All copper/zeolite samples possess a relatively high surface
area (>360 m2 g−1) with dispersion of 54–65% as tabulated in
Table 1. The copper loading is approximately 5 wt% for both
zeolites. CeO2 shows a relatively low surface area, due to low
porosity (<11 m2 g−1). The increase in weak acidity of Cu/HY
and Cu/HZSM-5, as compared to the parent zeolites (NH3-
TPD), is evidence for exchangeable copper cations in zeolites.
The Cu/HY and Cu/HZSM-5 samples show two reduction
peaks at 473 and 520 K (Fig. 1), corresponding to copper
oxide aggregates and highly dispersed copper oxide in the
pore of zeolite,30,31 respectively. The Cu dispersion on HY is
somewhat higher than that on HZSM-5 (peak at ~520 K), pre-
sumably due to a better diffusion of the Cu precursor in the
larger pore of HY. This is consistent with the TEM images
shown in Fig. 2. The Cu/HY possesses relatively small Cu par-
ticles (Fig. 2a) and some of them (dark spot) are well aligned
in the pore of HY (light grey plane), while a relatively lower
Cu dispersion on HZSM-5 can be evidenced by large semicir-
cle Cu particles, deposited on the external surface of HZSM-5
crystals (Fig. 2b).
measured
using
a
nitrogen
adsorption
analyzer
(Quantachrome) at 77 K and 0.05–0.30 P/P0. Reducible metal
oxide species in the catalysts were analyzed by temperature
programmed reduction (TPR). The catalysts were treated in
air at 723 K for 5 hours prior to heating from 323 to 1173 K
in 10% H2/Ar. The hydrogen consumption was recorded
using an on-line thermal conductivity detector (VICI).26 Cop-
per dispersion on a zeolite support was also analyzed by the
surface-selective TPR technique. Briefly after typical TPR, the
sample was in situ treated with N2O for selective oxidation of
surface copper to copperIJI) oxide at 333 K for 2 hours. Then,
the surface-oxidized sample was subjected to a secondary
TPR, in which the reduction took place only at the copperIJI)
oxide on the surface. The surface copper could be calculated
directly from hydrogen consumption of CuIJI) → Cu(0) as
described by Hoang et al.27 Cu dispersion is referred to as
the mole ratio of Cu on the surface over the bulk. The details
of calculations for Cu dispersion using this procedure can be
Table 1 % Dispersion, copper area, surface area, and acidity of copper catalysts and supports
Acidity (μmol g−1
)
% Cu
loading (wt%)
%
Cu area
BET surface
area (m2 g−1
Catalyst
Si/Al
Dispersion
(m2 gCu
)
)
Weak
Strong
CeO2
HY
HZSM-5
Cu/HY
Cu/HZSM-5
CeO2–Cu/HYa
CeO2–Cu/HZSM-5a
—
—
—
—
5.1
5.2
1.3
1.3
—
—
—
65
54
—
—
—
—
—
499
414
—
11
713
376
568
361
150
98
—
—
65
65
62
20
15
5
153
152
166
173
166
173
54
56
158
92
43
26
—
a
Values estimated from those of the parent catalysts.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
suggesting that the ketonization activity can be readily pro-
moted over the CeO2 component, while the hydrogenation–
dehydration of the acetone produced is not facilitated at this
temperature. This is probably due to competitive adsorption
by acetic acid over the Cu/HY at low temperature. It is worth
noting that, over Cu/zeolites, the hydrogenation–dehydration
of acetone alone can be accomplished at >473 K, as reported
in previous work.22 When the temperature was increased
from 548 to 598 K, the selectivity to propylene increases with
the decrease in acetone. This is not only because acetic acid
is less competitive at this temperature, but also the hydroge-
nation–dehydration is increasingly promoted by the Cu/HY
component. However, the propylene selectivity decreases with
a sharp increase in propane at >598 K despite that the Cu
alone cannot promote propylene hydrogenation.22,32 The
observed propane yield in this case is presumably derived
from the H-transfer process. As propylene is formed, it would
be protonated on the acid sites within a proximate vicinity of
Cu. Hence, some of the protonated propylene can be hydro-
genated by H-transfer from the proximate metal sites. This is
particularly the case at high temperature since H-transfer can
also be promoted from the hydrocarbon pools.33–35
Fig. 1 Temperature program reduction of 5% Cu/zeolites.
3.2 Catalytic activity testing
Effect of reaction temperature. Fig. 3 displays the effect of
temperature on acetic acid conversion over the three-
component catalyst containing 75 wt% of CeO2 and 25 wt%
of Cu/HY (CeO2–Cu/HYIJ25)). The conversion increases gradu-
ally with the reaction temperature and reaches 100% at above
598 K (Fig. 3a). The C3 products including acetone, propyl-
ene, and propane and CO2 are produced via ketonization of
acetic acid and subsequent hydrodeoxygenation of the ace-
tone formed, the so-called keto-hydrodeoxygenation (KHDO)
process. The selectivity to acetone is relatively high at 548 K
As mentioned earlier that acetic acid is competitively
adsorbed on Cu/HY at low temperature, the C2 products
including acetaldehyde, ethanol, ethylene, and ethane are
also observed, together with ethyl acetate (Fig. 3b). The result
from acetic acid conversion over Cu/zeolites alone (Table 2)
reveals that these C2 products are derived from direct hydro-
deoxygenation (HDO) of acetic acid. Over a copper catalyst,
Fig. 2 (a) TEM images of Cu/HY (i) at 440 kX, (ii) at 285 kX, and (iii) insertion (1,580 kX) with background/contrast adjustment to display the Cu
particle well aligned in the HY pore. (b) TEM images of Cu/HZSM-5 (i) at 440 kX and (ii) at 97 kX.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
ethanol can be readily dehydrated to ethylene over the acid
sites at high temperature, the acetaldehyde–ethanol equilib-
rium is not limited over Cu/HY as ethylene is formed. Accord-
ingly, acetaldehyde is largely consumed by the hydrogena-
tion–dehydration process as the temperature is raised. In a
manner similar to propane, ethane can also be produced by
H-transfer to ethylene at a relatively high temperature.
In fact, the observed high selectivity to ethyl acetate
(Fig. 3a) at low temperature suggests that ethanol is readily
formed but captured by acetic acid via esterification. How-
ever, the selectivity to ester decreases when the temperature
is increased due to the decreasing amount of acetic acid. It is
noted that no esterification between acetic acid and
i-propanol was found probably due to rapid dehydration of
the i-propanol, as compared to the ethanol. The selectivity to
ethanol, ethylene and ethane decreases at >598 K (Fig. 3b),
while C3 hydrocarbons are largely observed. This is because
the KHDO of acetic acid is more favorable at high tempera-
ture, as compared to the direct HDO. This is consistent with
the observed increase in ketonization activity of CeO2 at high
temperature as shown in Fig. S1.† It is worth noting that, as
the catalyst contains acidic zeolite (CeO2–Cu/HYIJ25)), the eth-
ylene and propylene produced can be further oligomerized to
C4–C5 olefins, particularly at >598 K. According to the prod-
ucts observed, the overall reaction scheme for keto-
hydrodeoxygenation of acetic acid can be proposed in Fig. 4.
Effect of the catalyst composition. The KHDO of acetic
acid depends largely on the component of the catalyst, as
shown by the experiments with various weight % of the CeO2
in the catalyst composition (Fig. 5). It is clear that the acetic
acid conversion increases when the CeO2 is increased
(Fig. 5a). Since the ketonization of acetic acid is primarily
promoted in KHDO, the catalyst with high CeO2 content
would be more active for overall acid conversion. Consistent
with this result, the yields of KHDO products, including ace-
tone (Fig. 5a) and propylene (Fig. 5b), are also increased,
together with CO2. However, the insufficient Cu/HY results in
a decrease in the subsequent HDO activity, as shown by a
decrease in propylene yield when CeO2 content is higher than
75 wt% (CeO2–Cu/HYIJ25), Fig. 5b). The increase in CeO2 plays
no significant role for direct HDO of acetic acid initially. The
yields of ethyl acetate and acetaldehyde remain similar from
Fig. 3 Effect of temperature on acetic acid KHDO over CeO2–Cu/
HYIJ25) (the number in parentheses represents wt% of Cu/zeolite in the
three-component catalyst) at 457 g h mol−, H2 30 ml min−1. a) Conver-
sion, propylene, propane, acetone, ethyl acetate, carbon dioxide, b)
ethylene, ethane, n-butane, C5 olefins, acetaldehyde, ethanol.
the acetic acid can be hydrogenated to acetaldehyde and then
ethanol. Acetaldehyde selectivity is decreased, while those of
ethanol, ethylene, and ethane are increased when the reac-
tion temperature is raised from 548 to 598 K (Fig. 3b). Since
Table 2 Direct HDO of acetic acid over 5% Cu/zeolitesa
Cu/HY
Cu/HZSM-5
Feed
Acetic acid
24.2
50 wt% aq. acetic acid
2.9
Acetic acid
13.9
Conversion (Cmol%)
Selectivity (Cmol%)
Acetaldehyde
Ethanol
Ethyl acetate
Ethylene
11.7
2.8
81.6
3.9
—
16.4
—
83.6
—
—
—
24.2
—
42.1
15.1
11.3
1.8
Acetone
Propylene
—
CO2
—
—
4.3
a
48 g h mol−1 and 573 K, H2 30 ml min−1, 1st hour on stream.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
the acetone in the zeolite and also the lower Cu dispersion as
mentioned earlier (Fig. 2).
Although CeO2–Cu/HZSM-5IJ25) seems to be less effective
for KHDO to produce propylene, this catalyst provides high
selectivity for direct HDO of acetic acid to ethylene. It can be
seen from Fig. 6c that a higher yield to ethylene can be
obtained from CeO2–Cu/HZSM-5IJ25). In the opposite manner,
CeO2–Cu/HYIJ25) gives mainly ethyl acetate that is an ester of
acetic acid and ethanol, derived from direct HDO of acetic
acid (Fig. 6a). In a support manner, the results in Table 2
provide the same conclusion as above. Over Cu/HY, high
selectivity to ethyl acetate is obtained, while ethylene selectiv-
ity is significantly enhanced over Cu/HZSM-5. This is presum-
ably because esterification can be inhibited by the confine-
ment of the medium pore zeolite; HZSM-5, which in turn,
promotes mainly monomolecular dehydration of the ethanol
formed to ethylene. It is worth emphasizing again that CeO2–
Fig. 4 The reaction network for acetic acid KHDO over CeO2–Cu/
zeolite.
60 to 75 wt% CeO2 (CeO2–Cu/HYIJ40) to CeO2–Cu/HYIJ25)). The
observed increase in ethanol yield (Fig. 5b) is due to a
reduced concentration of acetic acid that remained in the
reaction stream when ketonization is boosted. In line with
this view, dehydration to ethylene is also promoted initially.
However, at CeO2 content higher than 86 wt% (CeO2–Cu/
HYIJ14)), the overall HDO activity is significantly suppressed.
This is shown by the decrease in propylene, ethyl acetate, eth-
anol, acetaldehyde and ethylene. According to the results, it
is clear that CeO2 plays a significant role for initial activation
of acetic acid to acetone, while CuHY is essentially required
for olefin production from both HDO of acetone produced
and direct HDO of acetic acid. Hence, the combination of
these catalysts would lead to a successful olefin production
from acetic acid. Under the reaction conditions used in this
study, optimum yields of propylene and ethylene can be
obtained over CeO2–Cu/HYIJ25) with ~55% conversion.
Effect of the zeolite framework. It is clear from Fig. 6a
that CeO2–Cu/HZSM-5IJ25) is more active for acetic acid con-
version, as compared with CeO2–Cu/HYIJ25). This is because a
similar level of conversion is obtained from both catalysts,
despite that Cu/HZSM-5 possesses lower acidity (by NH3-TPD,
Table 1). However, at a similar level of conversion, the ace-
tone yield from CeO2–Cu/HZSM-5IJ25) (~34%) is higher than
that from CeO2–Cu/HYIJ25) (~24%), while the propylene yield
from both catalysts is not so significantly different (3–5% dif-
ferences) (Fig. 6b). At steady state, one may expect that as a
higher amount of acetone is left unconverted, the yield of
propylene should be accordingly lower. Nevertheless, the
observed increased yield of acetone without proportional
change in propylene yield can be attributed to the fact that
ketonization of acetic acid to acetone can also be facilitated
by HZSM-5IJ250).36,37 This is evidenced by a noticeable selec-
tivity to acetone (11.3%) when acetic acid was fed over Cu/
HZSM-5 alone (Table 2). Accordingly, the ketonization of
acetic acid is additionally promoted as seen by a higher yield
to CO2 over CeO2–Cu/HZSM-5IJ25). However, the further
hydrogenation–dehydration of the acetone produced to form
propylene is somewhat limited over this catalyst. This is
probably due to the competitive adsorption of acetic acid over
Fig. 5 Effect of CeO2–Cu/HY composition (from CeO2–Cu/HYIJ40) to
CeO2–Cu/HYIJ7)) on acetic acid KHDO at 457 g h mol−1 and 573 K, H2 30
ml min−1, 1st hour on stream. a) Conversion, acetone, ethyl acetate,
carbon dioxide. b) Propylene, i-propanol, ethylene, acetaldehyde, ethanol.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Cu/HZSM-5IJ25) provides much lower selectivity to ethyl ace-
tate despite that the same level of conversion is obtained for
both catalysts. This means that acetic acid must be converted
to acetone from both CeO2 and zeolite components in the
CeO2–Cu/HZSM-5IJ25). In a different manner, the ketonization
is promoted solely by the CeO2 component in the CeO2–Cu/
HYIJ25) and the Cu/zeolite component of this catalyst facili-
tates mainly hydrogenation–dehydration and also the
esterification.
Although several reaction networks take place over these
three component catalysts, both CeO2–Cu/HZSM-5IJ25) and
CeO2–Cu/HYIJ25) possess relatively high stability up to 60 and
30 hours on stream, respectively. However, a slight drop in
hydrogenation–dehydration over CeO2–Cu/HZSM-5IJ25) can be
noticed. This is presumably because ketonization of acetic
acid to acetone also takes place over the Cu/HZSM-5 compo-
nent, as discussed earlier. Hence, some of the acetone
formed over the acid site may well undergo aldol condensa-
tion to higher MW products deposited in the pores.38,39 On
the opposite manner, no ketonization of acetic acid takes
place over Cu/HY, and only hydrogenation–dehydration of
acetone produced over the CeO2 component is promoted over
the Cu component in CeO2–Cu/HYIJ25). Together with high
dispersion of the Cu active site in this catalyst, acetone can
be readily hydrogenated. Hence, the CeO2–Cu/HYIJ25) pro-
vides a relatively higher stability for propylene production, as
compared to the CeO2–Cu/HZSM-5IJ25).
Physical mixed bed vs. sequential bed. To minimize direct
HDO of acetic acid taking place over Cu/zeolite, two sepa-
rate beds containing CeO2 and Cu/HY catalysts were tested
at the same total contact time (343 + 114 g h mol−1) as the
physical mixed bed (457 g h mol−1). It can be seen from
Table 3 that the sequential bed gives a relatively higher con-
version, as compared to the physical mixed bed system. This
is because, in the physical mixed bed, part of acetic acid is
in contact with Cu/zeolite that possesses a higher surface
area, as compared to the CeO2. Hence, a relatively less frac-
tion of acetic acid is in contact with the CeO2 and a lower
activity can be expected over the physical mixed bed. This is
because the ketonization is kinetically second order and
highly sensitive to the partial pressure of acetic acid.25,40
Accordingly, as the virtual pressure of acetic acid on the cata-
lyst surface is reduced, the rate is markedly decreased. In a
support manner, the reaction with the same contact time,
but higher acetic acid partial pressure, provides a higher
activity (Table 3). This emphasizes the role of acetic acid
partial pressure on the ketonization activity, as discussed
previously.
For the sequential bed, the acetic acid is solely in contact
with CeO2 in the first bed. Accordingly, a higher conversion
of acetic acid to acetone can be expected, as also seen from
an increase in CO2 selectivity. The higher ketonization effi-
ciency in the sequential bed also leads to the higher propyl-
ene selectivity after the second bed (Cu/HY). This is because
the competitive adsorption by acetic acid over Cu/HY is
reduced as more acetone is produced from the first bed and
Fig.
6 Comparison between CeO2–Cu/HYIJ25) (open symbols) and
CeO2–Cu/HZSMIJ25) (filled symbols) for acetic acid KHDO at 457 g h
mol−1 and 573 K, H2 30 ml min−1. a) Conversion (circles), acetone
(triangles), ethyl acetate (squares). b) Propylene (circles), acetaldehyde
(triangles). c) Carbon dioxide (circles), ethylene (triangles).
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
Table 3 KHDO of acetic acid over CeO2–Cu/HYIJ25) with different systems and acetic acid concentrationsa
Mixed catalyst (457 g h mol−1
)
Sequential bed (343 + 114 g h mol−1
)
Acetic acid in H2 (mol%)
Conversion (Cmol%)
Selectivity (Cmol%)
Acetone
Propylene
Propane
1.2
55.6
2.3
85.4
1.2
88.9
27.2
16.4
—
8.8
45.9
2.6
1.7
49.4
7.8
Acetaldehyde
Ethanol
3.4
5.2
2.6
5.2
2.4
5.9
Ethyl acetate
Ethylene
29.6
2.5
7.7
3.6
4.3
4.1
Ethane
—
1.4
4.4
CO2
14.7
19.2
19.6
a
573 K, H2 30 ml min−1, 1st hour on stream.
less acetic acid remained in the reaction stream. Accordingly,
hydrodeoxygenation of acetone can be readily promoted,
while the products from direct HDO of acetic acid are
noticeably lower than those obtained from the mixed cata-
lyst (i.e. ethyl acetate selectivity decreases from 29.6 to
4.3%). However, the paraffinic hydrocarbons (propane and
ethane) were detected from the sequential bed system,
probably due to higher H-transfer efficiency at high
conversion.
not permanently deactivate the catalyst. As seen in Fig. 7, the
activity and selectivity of all products were recovered after
water was removed.
Light distillated hydrocarbons from KHDO of acetic acid.
In the practical point of view, a separate bed system would
provide more advantage in controlling and tuning the
Effect of water. Together with acetic acid, water is always
present in the gas stream from the pyrolysis biomass. Hence,
the effect of water on the direct HDO is primarily studied as
shown in Table 2. It can be seen that the conversion is signif-
icantly decreased when 50 wt% acetic acid (aq) is used as
feed. This is due to (i) the effect of feed dilution as men-
tioned earlier and (ii) an interaction of water with the three
component catalyst (CeO2–Cu/HYIJ25)). To verify the latter, an
experiment with alternate switching between acetic acid and
aqueous acetic acid was tested as displayed in Fig. 7. It can
be seen that the acetic acid conversion drops markedly after
the feed is replaced by aqueous acetic acid. Nevertheless, the
activity can be fully recovered after water withdrawal. This is
clearly due to the competitive adsorption between acetic acid
and water on the catalyst (Fig. 7a). The presence of water
strongly affects the olefin yields including propylene and eth-
ylene. This is not only because the conversion is decreased,
but water would also interact preferentially with the zeolites.
Hence, hydrogenation–dehydration of the acetone formed
and direct HDO of acetic acid are somewhat inhibited. The
observed similar yield of acetone after introducing water is
attributed to a combination effect between (i) a reduced keto-
nization (to form acetone) over the CeO2 component and (ii)
a decrease in hydrodeoxygenation (to convert acetone) over
the Cu/HY. The presence of water also decreases the yield of
ethyl acetate, as the esterification is a reversible process.
Moreover, the intermediate carbonyl compound, i.e. acetalde-
hyde, is increased with a decrease in ethanol, as the hydroge-
nation activity is suppressed (Fig. 7b). These results are in
agreement with the direct HDO of acetic acid over Cu/HY as
tabulated in Table 2. It is worth noting again that water does
Fig. 7 Effect of water on acetic acid KHDO over CeO2–Cu/HYIJ25) at
457 g h mol−1 and 573 K, H2 30 ml min−1. a) Conversion, acetone,
propylene, ethyl acetate. b) Ethylene, ethanol, acetaldehyde, carbon
dioxide.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
4. Conclusion
The acetic acid can be successfully converted to propylene via
keto-hydrodeoxygenation (KHDO) over the three component
catalyst containing CeO2 and Cu loaded acid zeolites (CeO2–
Cu/HZSM-5IJ25) or CeO2–Cu/HYIJ25)) at relatively low tempera-
ture (573 K) and atmospheric pressure. Ketonization of acetic
acid to acetone initially takes place and is followed by hydro-
genation–dehydration of the acetone formed to propylene.
Direct hydrodeoxygenation of acetic acid is also observed pro-
ducing ethylene and other C2-oxygenates. With higher Cu
dispersion, CeO2–Cu/HYIJ25) provides higher stability and
more selective conversion of acetic acid to olefin products, as
compared to that over CeO2–Cu/HZSM-5IJ25). The presence of
water suppresses the overall activity, particularly the hydroge-
nation–dehydration. With a sequential bed system composed
of CeO2, Cu/HY and HZSM-5, the KHDO can be applicable
for the conversion of acetic acid, a biomass derived product,
to olefins and higher hydrocarbons.
Acknowledgements
The authors are grateful for the financial support from the
Thailand Research Fund through the Royal Golden Jubilee
Ph.D. Program (Grant No. PHD/0137/2553) and for the assis-
tance from PTT Public Co. Ltd.
References
1 O. Kutsch, Market Study: Propylene, Ceresana, 2014.
2 M. Eramo, Global Ethylene Marketing Outlook [slide], IHS
Chemical, 2015.
3 C. A. Mullen and A. A. Boateng, Energy Fuels, 2008, 22,
2104–2109.
4 J. J. Bravo-Suarez, V. Schwartz and M. K. Kidder, Novel
Materials for Catalysis and Fuels Processing, American
Chemical Society, USA, 2013.
Fig. 8 Light distillated hydrocarbons from acetic acid over sequential
beds of CeO2–Cu/HY–HZSM-5 at 343–114–80 g h mol−1 and 573 K, H2
30 ml min−1. a) Conversion, propylene, acetone, ethylene, carbon
dioxide. b) Butenes, paraffins C2–C4, olefins C5–C10, aromatics C7–C9.
5 A. Seidel, Kirk-Othmer encyclopedia of chemical technology
vol. 1, John Wiley and Sons, Inc., New York, 1998.
6 G. A. H. Mekhemer, S. A. Halawy, M. A. Mohamed and M. I.
Zaki, J. Catal., 2005, 230, 109–122.
7 J. C. Kuriacose and R. Swaminathan, J. Catal., 1969, 14,
348–354.
8 R. Swaminathan and J. C. Kuriacose, J. Catal., 1970, 16,
357–362.
9 T. N. Pham, T. Sooknoi, S. P. Crossley and D. E. Resasco,
ACS Catal., 2013, 3, 2456–2473.
10 R. Martinez, M. C. Huff and M. A. Barteau, J. Catal.,
2004, 222, 404–409.
11 R. W. Snell and B. H. Shanks, Appl. Catal., A, 2013, 451,
86–93.
12 T. Waters, R. O'Hair and A. G. Wedd, Int. J. Mass Spectrom.,
2003, 228, 599–611.
13 J. J. Vajo, Y. K. Sun and W. H. Weinberg, J. Phys. Chem.,
1987, 91, 1153–1158.
catalyst performance. As olefins are obtained from KHDO
of acetic acid, adding another sequential bed of HZSM-5
(Si/Al ~ 13) can lead to a production of higher hydrocar-
bons as shown in Fig. 8. With excess catalyst (the sequen-
tial beds of CeO2–CuHY–HZSM-5), the hydrocarbon prod-
ucts including light olefins (ethylene, propylene, and
butylene), LPG (C2–C4 paraffins), gasoline (C5–C10 ole-
fins), and C7–C9 aromatics can be obtained from acetic
acid. A slight increase in selectivity to intermediate acetone
with a decrease in paraffin production indicates that the
deactivation also takes place, presumably over the Cu/HY
catalyst, while the continuous increase in propylene selec-
tivity with a decrease in C5–C10 olefins and C7–C9 aro-
matics selectivity is evidence for the HZSM-5 deactiva-
tion, which is typically observed for the acid-catalyzed
reaction. Nevertheless, this clearly demonstrates a potential
catalyst system/process for conversion of acetic acid to
hydrocarbon products.
14 S. Zhang, X. Duan, L. Ye, H. Lin, Z. Xie and Y. Yuan, Catal.
Today, 2013, 215, 260–266.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
15 J. Cressely, D. Farkhani, A. Deluzarchi and A. Kiennemann,
Mater. Chem. Phys., 1984, 11, 413–431.
16 V. J. Johnston, B. Kimmich, J. Potts, H. Weiner, J. H. Zink,
J. T. Chapman, L. Chen and R. Jevtic, US Pat., 0263910 A1,
2011.
27 D. L. Hoang, T. T. H. Dang, J. Engeldinger, M. Schneider, J.
Radnik, M. Richter and A. Martin, J. Solid State Chem.,
2011, 184, 1915–1923.
28 G. V. Sagar, P. V. R. Rao, C. S. Srikanth and K. V. R. Chary,
J. Phys. Chem. B, 2006, 110, 13881–13888.
17 Y. Zhu and X. W. L. Shi, Bull. Korean Chem. Soc., 2014, 35,
141–146.
29 A. Ausavasukhi, S. Suwannaran, J. Limtrakul and T. Sooknoi,
Appl. Catal., A, 2008, 345, 89–96.
18 W. Rachmady and M. A. Vannice, J. Catal., 2000, 192,
322–334.
19 H. Wan, R. V. Chaudhari and B. Subramaniam, Energy Fuels,
2013, 27, 487–493.
20 K. Zhang, F. Li, H. Zhang, H. Ma, W. Ying and D. Fang,
International Scholarly and Scientific Research & Innovation,
2013, vol. 7, pp. 488–492.
21 Y. K. Kharaica, W. W. Carothers and R. J. Rosenbauer,
Geochim. Cosmochim. Acta, 1982, 47, 397–402.
22 A. Witsuthammakul and T. Sooknoi, Catal. Sci. Technol.,
2015, 5, 3639–3648.
23 C. Liu, A. M. Karim, V. M. Lebarbier, D. Mei and Y. Wang,
Top. Catal., 2013, 56, 1782–1789.
24 R. W. Snell and B. H. Shanks, ACS Catal., 2014, 4, 512–518.
25 T. N. Pham, D. Shi and D. E. Resasco, Top. Catal., 2014, 57,
706–714.
30 Z. Strassberger, A. H. Alberts, M. J. Louwerse, S. Tanase and
G. Rothenberg, Green Chem., 2013, 15, 768–774.
31 C. Huo, J. Ouyang and H. Yang, Sci. Rep., 2014, 4, 1–9.
32 L. K. Frevel, M. J. Kressley and L. J. Kressley, US Pat.,
3076858, 1963.
33 S. Ilias and A. Bhan, ACS Catal., 2013, 3, 18–31.
34 N. Ravasio and M. Gargano, J. Org. Chem., 1993, 58, 1259–1261.
35 G. Parravano, J. Catal., 1970, 16, 1–15.
36 A. Gumidyala, T. Sooknoi and S. Crossley, 247th ACS
National Meeting & Exposition, 2014, p. 84.
37 A. Gumidyala, M. Godavarthy, T. Sooknoi and S. Crossley,
AIChE Annual Meeting, 2014.
38 J. Novakova and L. Kubelkova, J. Catal., 1990, 126, 689–692.
39 T. Tago, M. Sakamoto, K. Iwakai, H. Nishihara, S. R. Mukai,
T. Tanaka and T. Masuda, J. Chem. Eng. Jpn., 2009, 42,
s162–s167.
26 N. Peamaroon and T. Sooknoi, Pet. Sci. Technol., 2012, 30,
1647–1655.
40 T. N. Pham, D. Shi and D. E. Resasco, J. Catal., 2014, 314,
149–158.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.