Reaction network of aldehyde hydrogenation over sulfided Ni-Mo/Al 2O3 catalysts

Article abstract of DOI:10.1016/j.jcat.2004.12.010

A reaction network of aldehyde hydrogenation over NiMoS/Al ₂O₃ catalysts was studied with aldehydes with straight and branched carbon chains and different chain lengths as feed materials. The reactions in the gas phase and the liquid phase were compared. The main reaction in the aldehyde hydrogenation process is the hydrogenation of the CO double bond, which takes place over the coordinatively unsaturated sites. The major side reactions are self-condensation of aldehydes and condensation of aldehydes with alcohols. Both reactions involve α-hydrogen and are primarily catalyzed by acid-base bifunctional sites over the exposed Al₂O ₃ surfaces.

Full text of DOI:10.1016/j.jcat.2004.12.010

Journal of Catalysis 231 (2005) 20–32
www.elsevier.com/locate/jcat
Reaction network of aldehyde hydrogenation over
sulfided Ni–Mo/Al O catalysts
2
3
a
b
a,∗
Xueqin Wang , Ramzi Y. Saleh , Umit S. Ozkan
a
Department of Chemical Engineering, The Ohio State University, 140 West 19th Avenue, Columbus, OH 43210, USA
b
Global R&D-Products, ExxonMobil Chemical Company, Baton Rouge, LA 70821, USA
Received 19 October 2004; revised 5 December 2004; accepted 6 December 2004
Abstract
A reaction network of aldehyde hydrogenation over NiMoS/Al O catalysts was studied with aldehydes with straight and branched carbon
2
3
chains and different chain lengths as feed materials. The reactions in the gas phase and the liquid phase were compared. The main reaction
in the aldehyde hydrogenation process is the hydrogenation of the C=O double bond, which takes place over the coordinatively unsaturated
sites. The major side reactions are self-condensation of aldehydes and condensation of aldehydes with alcohols. Both reactions involve
α-hydrogen and are primarily catalyzed by acid–base bifunctional sites over the exposed Al O surfaces.
2
3

2004 Elsevier Inc. All rights reserved.
Keywords: Aldehyde hydrogenation; Sulfided Ni–Mo catalysts; Aldol condensation
1
. Introduction
genation (HDO), and hydrogenation (HYD). According to
a widely accepted model by Topsøe and co-workers, sulfi-
dation of the oxide phase results in the formation of stacks
of MoS2 slabs over the catalyst surface, and the Ni and Co
species are located primarily on the edges of these stacks
[4–7]. Active site models based on these structures have
been used to account for different product distributions ob-
served in HDS, HDN, and HDO reactions [8–18]. The im-
portance of surface SH or sulfhydryl groups has also been
discussed with regard to several reaction mechanisms [17,
19–23]. The evidence for the presence of SH groups has
been provided by deuterium exchange studies [24], chemi-
cal titration by silver ions [22], Raman spectroscopy [25],
and infrared spectroscopy [26]. The Brønsted acidity associ-
ated with SH groups in sulfided catalysts was also examined
by IR spectroscopy of adsorbed pyridine species [27]. Direct
correlations between the HDS activity and SH concentration
have also been reported previously [28–32]. The effect of
the changes in the surface concentration of Brønsted acid
sites and sulfur vacancies on the hydrogenolysis and hydro-
genation activities in HDN and HDO reactions have been
discussed in the literature [17,19]. Hydrocracking activity of
Oxo process alcohols are a major class of organic chemi-
cals [1]. The resulting products from the oxo process may
be aldehydes, primary alcohols, or a mixture of the two.
Most of these oxo aldehydes are hydrogenated to the cor-
responding oxo alcohols. A wide range of heterogeneous
catalysts is used in the industry to convert aldehydes to al-
cohols. Although the absence of C=C unsaturation makes
the hydrogenation of oxo aldehydes easier, the presence of
ppm levels of sulfur in the feed olefin can place serious con-
straints on the catalyst choice. Sulfided Ni–Mo catalysts,
which are known for their hydrogenation and hydrogenol-
ysis activity [2,3], can hydrogenate aldehydes while main-
taining their activity in the presence of sulfur compounds,
which may be part of the feed stream. Sulfided Ni–Mo and
Co–Mo catalysts supported on Al2O3 are used extensively
in many hydrotreating processes, including hydrodesulfur-
ization (HDS), hydrodenitrogenation (HDN), hydrodeoxy-
*
Corresponding author. Fax: 614-292-9615.
E-mail address: ozkan.1@osu.edu (U.S. Ozkan).
0021-9517/$ – see front matter  2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2004.12.010

X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
21
the sulfhydryl groups was also reported [33,34]. It has also
been proposed that reduced Ni–Mo/Al2O3 catalysts may ex-
hibit similar activity, where OH groups act in a manner anal-
ogous to that of SH groups, and anion vacancies (CUS) can
form during the reduction process as well [19,35,36].
Although the relationships between different active sites,
such as coordinatively unsaturated sites (CUS), Brønsted
acid sites, and SH and OH groups, over NiMoS and CoMoS
catalysts and different reaction steps in HDS, HDN, and
HDO reactions are studied extensively, similar correlations
in aldehyde hydrogenation reactions have not been the fo-
cus of many studies. In our previous articles [37–39], we
have reported on the catalytic performance of sulfided Ni–
Mo/Al2O3 catalysts in the hydrogenation of linear aldehydes
to alcohols. The two primary reactions that are of most in-
terest in these reaction schemes are the hydrogenation of the
aldehyde to form a corresponding alcohol (the desired re-
action) and the formation of heavy products (the undesired
reaction). We have also shown that NO and CO2 could be
used as molecules to probe the active sites that promote the
hydrogenation and heavy product formation reactions, re-
spectively, as the alcohol formation is correlated with CUS
density, whereas the heavy product formation can best be ex-
plained by the surface concentration of OH and possibly SH
groups.
molecules as feed into the reactor system by saturating a
H2 stream with the use of temperature-controlled bubblers.
The concentration of the feed was controlled by the temper-
ature of the bubbler and verified by GC analysis. The outlet
of the reactor was connected directly to a condenser with
decane as solvent, which was cooled in an ice-water bath.
The outlet flow was switched to the condenser after the reac-
tion reached steady state. Condensed species were analyzed
by gas chromatography with the use of the liquid auto in-
jector. The lighter components were analyzed on line with
the gas injection mode of the GC. The liquid-phase reaction
was performed with the Fast Hydrogenation Autoclave Unit
(FHAU). The unit consists of two CSTRs, each equipped
with dual liquid (oil and water) and hydrogen feed systems,
a Robinson–Mahoney stationary catalyst basket, and a heat-
ing mantle. The unit was operated in the continuous mode,
where only one reactor train was used. The catalyst basket is
3
charged with 6 cm of catalyst. Under each set of conditions,
the reactor effluent was collected over a period of 4–6 h and
analyzed by an off-line GC equipped with a 60-m boiling
point column and FID. All by-products generated from gas-
phase and liquid-phase reactions were identified by GC/MS
and GC/IR.
TPD experiments were performed with a homemade ap-
paratus, which was previously described [42]. The reac-
tor effluent composition was continually monitored as a
function of sample temperature with a mass spectrometer
The focus of this study has been the reaction path-
way(s) involved in aldehyde hydrogenation and the rela-
tionship between reactions and nature of different surface
sites over NiMoS/Al2O3 catalysts. Aldehydes with straight
and branched carbon chains and different chain lengths were
studied as feed materials. The reactions in the gas phase and
the liquid phase were compared. Propanal has been used as
a model aldehyde for a more detailed analysis of reaction
intermediates, products, and side reactions.
(
Hewlett–Packard, MS Engine 5898A). For each of the TPD
2
runs, 30 m (surface area) of sample was loaded into the
U-tube quartz reactor. All samples were sulfided in situ by
the same procedure as in the reaction studies, that is, 10%
◦
H2S in H2 was passed over the sample for 10 h at 400 C,
followed by He flushing for 2 h at the same temperature
◦
and cooling to 30 C in He. Propanal adsorption was per-
formed by the introduction of propanal through a diffusion
3
◦
tube with a 30 cm (STP)/min He flow at 30 C for 1 h.
After being flushed with He for 1 h at the same tempera-
2
. Experimental
◦
ture, the samples were heated at a rate of 10 C/min under a
3
3
0 cm (STP)/min He flow.
Alumina-supported catalysts with different Mo and Ni
The conversion of aldehyde (C%) is calculated with the
equation
loadings were prepared by wet co-impregnation of γ -Al2O3
with aqueous solutions of ammonium heptamolybdate and
nickel nitrate. The preparation procedure has been presented
previously [37,40,41]. The catalyst compositions are re-
ported as weight percentages of the oxide precursors, that is,
MoO3 and NiO, following the convention commonly used in
literature. The surface area of samples used in these studies
varied between 166 and 195 m /g, with pure alumina giv-
ing the highest surface area. Before all reaction studies, the
catalysts were sulfided in situ at 400 C with 10% H2S in H2
moles aldehyde in feed − moles aldehyde in product
C% =
moles aldehyde in feed
×
100.
The selectivity of product i is defined as
moles of C in product i
2
Si% =
× 100.
moles of C in aldehyde converted
◦
for 10 h followed by He flushing for 1 h at the same temper-
ature before the system was cooled to the desired reaction
temperature.
3
. Results
The reaction studies were performed in both the gas phase
and the liquid phase. The gas-phase reaction was carried out
in a fixed-bed reactor system, which is described in a pre-
vious article [37]. We introduced propanal and other probe
3.1. Propanal hydrogenation
Blank reactor runs showed no detectable conversion in
the hydrogenation of propanal through an empty reactor

22
X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
Fig. 1. Conversion and selectivity in propanal hydrogenation over sul-
fided 3% Ni–15% Mo/Al O catalysts (1000 psi, 0.13% propanal in
2
3
3
◦
50 cm /min H ): (2, 1) 180, (", !) 160, and (Q, P) 140 C. Bold and
2
2
blank symbols represent propanal conversion and selectivity to propanol,
respectively.
(
without catalyst). As can be seen in Fig. 1, where con-
version and alcohol selectivity are presented for the first
0 h, reaction over sulfided 3% Ni–15% Mo/Al2O3 cata-
1
lyst reached steady state within the first 5 h. All data pre-
sented in this article were collected after steady state was
reached.
As expected, propanal conversion increases with in-
creasing reaction temperature. Meanwhile, selectivity for
propanol also increases with reaction temperature. How-
◦
ever, the difference between 180 and 160 C is much smaller
◦
compared with the difference between 160 and 140 C. Un-
der the experimental conditions used, the main product is
propanol, formed from hydrogenation of the C=O double
bond of propanal. Following the definition of “lights and
heavies” in hexanal hydrogenation used previously [37],
in this article “lights” refers to products with carbon num-
bers lower than that of the feed, which are produced mainly
from full hydrogenation, decarbonylation, and hydrogenol-
ysis/cracking. The term “heavies” refers to products heavier
than primary product alcohol, which are produced mainly
from self-condensation of aldehydes and condensation of
aldehyde with alcohol.
Fig. 2. Effect of feed concentration on propanal hydrogenation performance
3
over sulfided 3% Ni–15% Mo/Al2O3 catalysts (1000 psi, 250 cm /min
◦
H ): (a) selectivities at 180 C, (b) propanal conversion rates at different
2
temperatures, and (c) propanol formation rates at different temperatures.
In the process of high-pressure hydrogenation of alde-
hyde, reaction performance strongly depends not only on
the reaction temperature, but also on the feed concentration.
From Fig. 2a, it can be seen that selectivity for propanol de-
creases with feed concentration, whereas the selectivity for
heavies increases, suggesting that the formation of heavy
products strongly relies on feed concentration. Selectivity
for lights, the minor by-products, decreases slightly with
increasing feed concentration. Fig. 2b shows propanal con-
version rates to be proportional to feed concentration for all
temperatures tested. Propanol formation rates, however, do
not show a linear dependence on concentration (Fig. 2c), im-
plying that the primary product, propanol, is consumed in
secondary reactions.
3.2. Effect of structure and chain length of the aldehyde
molecule in the hydrogenation reaction
To gain insight into the mechanistic steps involved in
aldehyde hydrogenation reactions, four different aldehyde
molecules were tested: propanal, hexanal, 2-ethyl-butanal,
and 2-methyl-pentanal. The objective was to examine the ef-
fect of chain length and to compare the linear aldehydes to
branched aldehydes in hydrogenation reactions over sulfided
Mo/γ -Al2O3 and Ni–Mo/γ -Al2O3 catalysts.
3.2.1. Hydrogenation of propanal and hexanal
Fig. 3 shows the catalytic performance of sulfided Mo/
γ -Al2O3 and Ni–Mo/γ -Al2O3 catalysts in the hydrogena-

X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
23
◦
Fig. 3. Effect of catalyst composition (% NiO–% MoO /Al O ) on performance in hexanal and propanal hydrogenation at 180 C: (1) hexanal (0.09%),
3
2 3
(2) propanal (0.26%). (a) Conversion, (b) selectivity to alcohol, (c) selectivity to lights, and (d) selectivity to heavies.
tion of propanal and hexanal at the same reaction tempera-
ture of 180 C and a reaction pressure of 1000 psi. To make
3.2.2. Effect of temperature on the hydrogenation of
hexanal and 2-ethyl-butanal
◦
a comparison of product distributions at similar conversion
levels for the two feed molecules, we obtained the data in
these figures at propanal and hexanal concentrations of 0.26
and 0.09%, respectively (Fig. 3a).
The effect of temperature on selectivities for two dif-
ferent aldehyde molecules, hexanal and 2-ethyl-butanal, is
presented in Fig. 4. The two molecules chosen have the
same carbon number, but one is linear and the other has a
branched structure. Experiments were carried out over sul-
fided 3% Ni–15% Mo/γ -Al2O3 catalysts at a temperature
Both propanal and hexanal conversions and selectivities
for alcohols increase with Mo loading over mono-metallic
catalysts. Although Ni addition increases the conversion,
further increases in Ni content do not change conversion
or alcohol selectivity very noticeably. For both aldehydes,
the selectivity for lights increases with increasing Mo con-
tent, and the selectivity for heavies decreases. The effect of
Ni content is much more pronounced for light product se-
lectivity than it is for heavy product selectivity. Although
the trends for the two aldehyde molecules are quite simi-
lar, the alcohol selectivity for hexanal is significantly higher
than it is for propanal. Another major difference between
the two feed molecules is that the selectivity for heavies for
propanal is much higher, whereas the selectivity for lights is
lower. It is conceivable that the longer chains can undergo
cracking more readily, giving rise to more lights production.
The higher heavy product selectivities observed for propanal
hydrogenation imply that smaller aldehydes can undergo
condensation-type side reactions more easily. However, the
difference in heavy selectivities can also be explained by the
different feed concentrations, as mentioned in the previous
section.
◦
range of 140–180 C and 1000 psi H2. The aldehyde con-
centration was kept at 0.8–0.9%. Although the conversions
of the two aldehyde feed molecules are comparable, the se-
lectivity trends observed are not very similar. The alcohol
selectivity increases with temperature for hexanal, whereas
it is seen to decrease for 2-ethyl-butanal. The selectivities
for light and heavy products are also quite different for the
branched and linear aldehydes. The selectivity for lights is
much higher for the branched molecule, with a sharp in-
crease with increasing temperature. The same selectivity for
the linear molecule, on the other hand, is much lower and
shows only a small increase with increasing temperature.
The trends for the heavy products are exactly the opposite of
those seen for the light products. The linear aldehyde shows
a much higher selectivity for heavy products, which de-
creases sharply with increasing temperature. The branched
molecule, on the other hand, shows very little heavy product
formation, with little temperature sensitivity. The calculated
apparent activation energies are 11 and 9 kcal/mol for the
hydrogenation of hexanal and 2-ethyl-butanal, respectively.

24
X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
Fig. 4. Effect of temperature on performance of hexanal and 2-ethyl-butanal hydrogenation over sulfided 3% Ni–15% MoO /Al O catalysts. (a) Conversion
3
2 3
of aldehyde, (b) selectivity to alcohol, (c) selectivity to lights, and (d) selectivity to heavies; (") hexanal (0.09%), (Q) 2-ethyl-butanal (0.08%).
Table 1
length and branched structure on the conversion of aldehyde
and selectivity for by-products are as follows:
Comparison of hydrogenation of propanal, hexanal, 2-ethyl-butanal, and
a
2-methyl-pentanal over sulfided 3% Ni–15% Mo/Al O catalysts
2
3
b
c
Feed molecule
C %
S %
• Conversion of aldehyde:
Aldehyde
Alcohol
Lights
Heavies
propanal > hexanal > 2-ethyl-butanal > 2-methyl-
pentanal.
Propanal
Hexanal
99.7
94.2
82.6
58.9
78.4
87.3
67.1
86.2
4.0
4.6
31.8
8.4
17.6
8.1
1.1
5.4
•
Selectivity for heavies:
propanal > hexanal > 2-methyl-pentanal > 2-ethyl-
2-Ethyl-butanal
2-Methyl-pentanal
butanal.
a
◦
3
Reaction conditions: T = 180 C, 1000 psi, 250 cm (STP)/min H ;
2
• Selectivity for lights:
concentration of aldehyde: propanal = 0.13%, hexanal = 0.09%, 2-ethyl-
propanal < hexanal < 2-methyl-pentanal < 2-ethyl-
butanal = 0.08%, 2-methyl-pentanal = 0.10%.
butanal.
b
C%—conversion.
S%—selectivity.
c
3
.3. Comparison of reactions in liquid phase and gas phase
For the branched molecule, cracking reactions are likely to
play a more important role in the overall conversion of the
aldehyde compared with the linear molecule. Therefore, the
difference in the activation energies does not necessarily im-
ply a difference in the intrinsic activation energy of the C=O
hydrogenation reaction, but it may simply be due to a contri-
bution from the cracking reaction for the branched molecule.
Table 1 shows a comparison of four different aldehyde
molecules, propanal, hexanal, 2-methyl-pentanal, and 2-eth-
yl-butanal, under similar conditions. The effect of chain
Liquid-phase reaction of hexanal hydrogenation was per-
formed in a CSTR unit. The products obtained from liquid-
phase reaction and gas-phase reaction (condensed in decane)
were analyzed by gas chromatography with a 60-m boil-
ing point column and FID. The sample chromatograms from
liquid-phase and gas-phase reaction experiments were very
similar, indicating that the products obtained from the gas-
phase and liquid-phase reaction studies are the same. De-
tailed identification of the product species was achieved by

X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
25
GC/MS and GC/IR analyses. The main by-products con-
sisted of acetals, dimer aldehydes and alcohols, dimer ethers,
esters, acids, and oxygenated trimers and tetramers.
are presented in Fig. 5b. The space velocities for both liquid-
phase and gas-phase reaction experiments are expressed as
LHSV for purposes of comparison. Alcohol selectivity in-
creases with increasing temperature for both liquid phase
and gas phase. It is also interesting to note that the selec-
tivities obtained at the same temperature are very close to
one another for both liquid and gas phases when space ve-
locities are comparable. The similarities between the trends
observed in the gas phase and the liquid phase are reassur-
ing, since they imply that the experimental results obtained
in the gas phase are still relevant for industrial applications
of similar reactions, which use primarily liquid feeds.
The effect of water addition was also examined in both
liquid-phase and gas-phase reactions. Table 2 clearly shows
that under identical conditions, higher alcohol selectivities
are obtained when water is introduced into the feed. This
observation holds for both liquid and gas phases and for
both hexanal and propanal feeds. It is possible that water,
by acting as a base, may be neutralizing the OH groups
on the Al2O3 surface that are mainly responsible for the
heavy product formation. The effect of H2O addition on the
main by-products in the liquid-phase reaction, identified by
GC/MS and GC/IR, is presented in Fig. 6. The heavies that
were significantly reduced with water addition are dimers
The effect of reaction temperature at different space ve-
locities on the performance of hexanal hydrogenation in the
liquid phase is shown in Fig. 5. The conversion increases al-
most linearly with temperature at the highest space velocity
used. This trend is not as apparent at lower space velocities,
since conversion approaches 100% at higher temperatures.
The apparent activation energy for liquid-phase reactions is
around 7 kcal/mol. The fact that it is somewhat lower than
the activation energy obtained in the gas phase raises the
question about possible mass transfer limitations in the liq-
uid phase. Alcohol selectivities for different space velocities
(C12 aldehydes and alcohols), esters, trimers, and tetramers.
Fig. 5. Effect of temperature and space velocity on hexanal conversion (a)
and selectivity to hexanol (b) in hexanal hydrogenation at liquid phase over
−
1
sulfided 3% Ni–15% MoO /Al O catalyst: (F) SV (12 h ), (2) SV
3
2 3
− −1
3.8 h ), and (Q) SV (2.0 h ); gas phase results are included for com-
1
Fig. 6. Effect of H O in the feed on formation of by-products in hexanal
2
(
−
1
hydrogenation at liquid phase over sulfided 3% Mo–15% Mo/Al O cata-
parison ((!) SV (0.8 h )). (a) Conversion of aldehyde, (b) selectivity to
2
3
lysts.
alcohol.
Table 2
Effect of H O on hydrogenation of hexanal and propanal (sulfide 3% Ni–15% Mo/Al O catalyst)
2
2 3
a
b
c
d
Condition
Hexanal liquid phase
Hexanal gas phase
Propanal gas phase
C%
S%
C%
S%
C%
S%
W/O H O
W/ H O
2
97.6
97.5
91.3
93.9
98.5
94.7
88.3
94.7
53.7
51.6
55.6
57.4
2
a
◦
Reaction temperature 180 C, H pressure 1000 psi.
2
b
c
d
−1
Space velocity 2.0 h , concentration of hexanal and H O, 20 and 0.60%.
Space velocity 1.2 h , concentration of hexanal and H O, 0.52 and 0.24%.
Space velocity 0.4 h , concentration of propanal and H O, 0.25 and 0.02%.
2
−1
2
−1
2

26
X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
Table 3
Product distribution of propanal hydrogenation over sulfided 7% Ni–
Table 4
a
Reaction of propanol over sulfided 7% Ni–15% Mo/Al O
2
3
a
1
5% Mo/Al O catalyst
2 3
Product
Selectivity of products (%)
Product
Selectivity (%)
◦
◦
◦
140 C
160 C
180 C
◦
◦
◦
1
40 C
160 C
180 C
Ethane
Propane
Iso-propanol
Methyl-propyl ether
Propyl ether
3.9
54.1
–
9.0
33.1
3.7
53.2
–
9.4
33.4
4.4
52.0
1.7
14.1
27.9
Ethane
Propane
Propanol
–
0.3
29.2
–
4.4
4.6
0.4
0.6
49.1
–
0.1
1.0
38.6
0.2
3.4
5.2
0.7
0.5
39.8
–
0.4
2.4
57.4
1.2
1.1
3.1
2
2
2
-Methyl-pentane
-Methyl-pentene
-Methyl-2-pentene
Conversion of propanol (%)
a
0.3
3.0
8.0
Propyl ether
Propionic acid
2.0
0.3
3
Reaction conditions: 1000 psi, 0.05% propanol in 250 cm (STP)/min
2
H , 12.5 m catalyst.
2
2
-Methyl-pentanal
Propyl-propionate
-Methyl-2-pentenal
Propionaldehyde dipropyl acetal
13.4
0.5
0.03
0.03
12.5
0.1
2
5.6
–
0.2
–
of alcohol was seen to be very low (Table 4). The main prod-
ucts are propane from hydrogenation of propanol and propyl
ether from bimolecular dehydration of two propanol mole-
cules. The trace amounts of ethane and methyl-propyl ether
observed are likely to result from cracking of propyl ether.
At higher temperatures, isopropanol appears because of the
isomerization of propanol, possibly catalyzed by acidic sites
2
2
-Methyl-pentanol
-Methyl-pentenol
0.7
0.4
5.2
5.2
0.2
5.3
Trimers
5.3
Conversion of propanal (%)
a
13.6
25.4
53.7
3
Reaction conditions: 1000 psi, 0.35% propanal, 250 cm (STP)/min
2
H , 12.5 m catalyst.
2
(
–OH, –SH). The comparison presented in Table 5, however,
shows that the effect of propanol on propanal hydrogenation
cannot be ignored. The presence of alcohol in the feed causes
a significant decrease in aldehyde conversion, suggesting a
strong inhibition effect. The product distributions in this ta-
ble are presented as flow rates since the presence of alcohol
in the feed makes the use of a “selectivity” concept difficult.
The amounts of 2-methyl-pentanal and trimers increase
significantly because of the enhancement of the reaction
between propanal and propanol. The 2-methyl-2-pentenal
and 2-methyl-2-pentenol yields decrease, possibly because
of the inhibition effect from adsorbed propanol on the self-
condensation reaction of propanals.
The addition of water, however, appears to produce more
acids. It is also noted that the presence of water in the feed
results in a decrease in the overall aldehyde conversion, since
H2O may inhibit the vacancy sites by adsorption and/or oxi-
dation.
3
.4. Reaction studies using intermediate and product
species as probe molecules
Since our previous experiments showed that propanal
and hexanal exhibit the same trends in hydrogenation re-
actions, further studies to elucidate the reaction networks
were conducted with C3 aldehydes and their intermediates
or products. Propanal, which can be representative of the
straight-chain aldehydes, better facilitates an examination of
the effect of intermediate species through easier analysis and
identification of all products. Reactions of probe molecules
were studied over a sulfided bimetallic catalyst, 7% Ni–
3
.4.3. Reaction of 2-methyl-pentanal
-Methyl-pentanal is one of the major by-products of
2
propanal hydrogenation. Table 6 shows the reaction results
obtained when 2-methyl-pentanal is used as a feedstock over
◦
sulfided 7% Ni–15% Mo/γ -Al2O3 catalyst. At 180 C, the
conversion of 2-methyl-pentanal is quite high, and the main
reaction is hydrogenation of aldehyde to the corresponding
alcohol. 2-Methyl-pentane and 2-methyl-pentyl ether, which
are seen in small amounts, result from the hydrogenation
of 2-methyl-pentanol and dehydration of 2-methyl-pentanol,
respectively. Propane may come from cracking of 2-methyl-
pentanal and 2-methyl-pentanol molecules.
1
5% Mo/γ -Al2O3.
3
.4.1. Propanal hydrogenation
Table 3 shows the product distribution of propanal hydro-
genation over sulfided 7% Ni–15% Mo/γ -Al2O3 catalyst at
a temperature range of 140–180 C. The main products from
◦
propanal hydrogenation at low reaction temperatures are
propanol, 2-methyl-pentanal, 2-methyl-pentenal, methyl-
pentenes, and trimers. Major products at high temperature
are propanol, 2-methyl-pentanal, 2-methyl-pentanol, and hy-
drocarbons. No significant change in selectivity for trimers
is observed.
3
.4.4. Reaction of 2-methyl-pentanol
2-Methyl-pentanol is the other major by-product of the
propanal hydrogenation process, especially at higher reac-
tion temperatures. When it is used as the feed molecule over
sulfided 7% Ni–15% Mo/γ -Al2O3 catalyst, the only major
reaction observed is the hydrogenation of alcohol to produce
the corresponding hydrocarbon, 2-methyl-pentane (Table 7).
Other products are propane and trace quantities of ethers.
3
.4.2. Effect of propanol on propanal hydrogenation
When the reactivity of propanol was examined under the
conditions used for propanal hydrogenation, the conversion

X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
27
Table 5
Product flow rates (mol/h × 10 ) in propanal hydrogenation (w/ and w/o propanol) over sulfided 7% Ni–15% Mo/Al O catalyst
7
a
2
3
Product
Without propanol
With propanol
◦
◦
◦
◦
◦
◦
1
40 C
160 C
180 C
140 C
160 C
180 C
Ethane
Propane
Propanol
–
0.1
0.9
38.3
0.1
1.8
2.5
0.4
0.4
19.7
–
0.1
–
2.5
0.1
1.7
1.2
5.1
120.3
1.2
1.2
3.3
–
0.1
0.8
52.3
–
0.6
4.3
116.0
0.6
1.8
4.3
1.0
0.8
20.7
0.6
–
0.2
0.4
42.6
–
15.6
–
2-Methyl-pentane
2-Methyl-pentene
2-Methyl-2-pentene
0.8
1.2
0.2
0.4
13.1
–
1.6
–
0.2
0.1
0.9
1.4
1.7
0.3
0.8
16.0
0.4
1.4
1.4
1.4
–
1.8
2.3
0.5
0.4
21.1
0.2
trace
0.1
2.5
–
Propyl ether
Propionic acid
2.2
0.8
2
-Methyl-pentanal
Propyl-propionate
-Methyl-2-pentenal
Propionaldehyde dipropyl acetal
14.1
0.4
trace
trace
13.1
0.1
2
0.1
11.3
–
2
2
-Methyl-pentanol
-Methyl-pentenol
Trimers
3.7
2.5
3.8
6.1
Conversion of propanal (%)
a
13.6
25.4
53.7
9.8
−7
14.5
39.0
3
2
Reaction conditions: 1000 psi, 250 cm (STP)/min H , 12.5 m catalyst, propanal flow rate in feed = 391 × 10 mol/h, propanol flow rate in feed =
2
6 × 10 7 mol/h. Trace means < 0.1 in the table.
−
5
Table 6
Reaction of 2-methyl-pentanal over sulfided 7% Ni–15% Mo/Al O cata-
lyst
Table 8
Reaction of 2-methyl-pentenal over sulfided 7% Ni–15% Mo/γ -Al O cat-
alyst
2
3
2 3
a
a
Product
Propane
Selectivity (%)
Product
Selectivity of products (%)
◦
◦
◦
2.0
3.1
90.3
4.7
140 C
160 C
180 C
2-Methyl-pentane
2-Methyl-pentanol
2-Methyl-pentyl ether
Propane
0.1
0.3
7.2
5.2
54.4
–
0.1
0.7
7.9
7.1
54.7
–
0.1
2.0
8.1
8.9
59.3
0.5
19.5
0.8
1.3
2-Methyl-pentane
2-Methyl-pentene
2-Methyl-2-pentene
2-Methyl-pentanal
Conversion of propanal (%)
a
62.4
◦
Reaction conditions: 180 C, 1000 psi, 0.05% 2-methyl-pentanal in
Propyl-propionate
2-Methyl-pentanol
3
2
2
50 cm (STP)/min H , 25 m catalyst.
2
30.3
1.2
1.2
27.3
1.0
1.2
2-Methyl-pentenol
Others
Table 7
Reaction of 2-methyl-pentanol over sulfided 7% Ni–15% Mo/Al O cata-
2
3
Conversion of propanal (%)
a
56.1
63.5
89.5
a
lyst
3
Reaction conditions: 1000 psi, 0.11% 2-methyl-pentenal in 250 cm
Product
Propane
Selectivity (%)
2
(STP)/min H , 25 m catalyst.
2
9.9
89.1
0.4
2
4
2
-Methyl-pentane
-Methyl-pentyl propyl ether
-Methyl-pentyl ether
3
.4.6. Reaction of acetal
0.6
A trace amount of propionaldehyde dipropyl acetal was
Conversion of propanal (%)
36.2
observed in the propanal hydrogenation reaction. Since pro-
pionaldehyde dipropyl acetal was not available, the reaction
experiments were performed with propionaldehyde diethyl
acetal to examine the reactivity of these acetals over the
sulfided Ni–Mo catalysts used in this study. The reaction
results, which are summarized in Table 9, show high reac-
tivity. The main products are ethanol and ethers, which are
produced from hydrogenolysis and cracking of acetal.
a
◦
Reaction conditions: 180 C, 1000 psi, 0.02% 2-methyl-pentanol in
3
2
2
50 cm (STP)/min H , 25 m catalyst.
2
3
.4.5. Reaction of 2-methyl-pentenal
-Methyl-pentenal is one of the major by-products at
2
lower temperatures in the propanal hydrogenation network.
Results from the reaction of 2-methyl-pentenal over sul-
fided 7% Ni–15% Mo/γ -Al2O3 catalyst are presented in
Table 8. 2-Methyl-pentanal and 2-methyl-pentanol are the
main products in this reaction. In addition, hydrocarbons,
such as 2-methyl-pentane, 2-methyl-pentene, and 2-methyl-
3.4.7. Reaction of propyl ether
A small amount of propyl ether was observed in propanal
hydrogenation. Reaction of ether as a feed over the same cat-
alyst showed a moderate conversion, and the main products
were propane and propanol (Table 10).
2
-pentene, are also produced. 2-Methyl-pentenol is also ob-
served.

28
X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
Table 9
Reaction of propionaldehyde diethyl-acetal over sulfided 7% Ni–15%
Table 11
a
Reaction of propanal over sulfided Al O support
2
3
a
MoO /Al O catalyst
3
2 3
Product
Selectivity of products (%)
Product
Selectivity (%)
◦
◦
◦
1
40 C
160 C
180 C
◦
◦
1
40 C
180 C
Propanol
Propyl ether
2-Methyl-pentanal
Propyl-propionate
2-Methyl-2-pentenal
Trimers
0.5
0.7
0.3
0.9
0.6
96.3
1.3
1.1
0.4
1.3
0.5
95.2
1.5
Ethane
0.8
0.5
35.8
0.3
0.6
0.3
0.3
7.2
32.5
14.9
7.6
4.1
5.0
40.9
0.3
6.4
0.7
5.2
17.0
9.0
3.1
8.5
–
–
Propane
Ethanol
Propanal
Iso-propanol
Propanol
0.6
97.8
1.1
Conversion of propanal (%)
a
8.8
18.9
27.8
Methyl ether
Methyl-ethyl ether
Ethyl-propyl ether
Allyl ethyl ether
Others
3
Conditions: 1000 psi, 0.35% propanal in 250 cm (STP)/min H ,
2
2
1
2.5 m γ -Al O .
2
3
Table 12
Conversion of acetal (%)
a
47.5
94.7
7
Product flow rates (mol/h ×10 ) in propanal reaction (w/ and w/o
a
Reaction conditions: 1000 psi, 0.02% propionaldehyde diethyl acetal,
propanol) over sulfided Al2O3 support
3
2
2
50 cm (STP)/min H , 12.5 m catalyst.
2
Product
Without propanol
With propanol
Lights
Iso-propanol
Propanol
Cycle-propane
Ethers
Propionic acid
–
–
1.2
–
0.2
–
0.7
0.3
51.7
–
1.2
0.8
Table 10
a
30.7
12.9
18.4
13.3
10.2
8.2
6.8
2.2
2.3
Reaction of propyl-ether over sulfided 7% Ni–15% Mo/Al O catalyst
2
3
Product
Selectivity (%)
◦
◦
1
40 C
180 C
2
-Methyl-pentanal
Propyl-propionate
-Methyl-2-pentenal
Ethane
Propane
Propanal
4.9
48.7
–
2.3
72.0
0.1
2
Propionaldehyde dipropyl acetal
Trimers
Iso-propanol
Propanol
Methyl ether
Methyl-propyl ether
Ethyl-propyl ether
Others
8.3
12.0
11.7
5.1
4.7
4.7
1.1
14.8
3.7
4.5
1.1
0.5
Conversion of propanol (%)
Conversion of propanal (%)
44.9
26.0
27.8
a
3
2
0.5
Reaction conditions: 1000 psi, 250 cm (STP)/min H2, 12.5 m
−
7
γ -Al O , propanal flow rate in feed = 391 × 10 mol/h, propanol flow
2
3
Conversion of propyl-ether (%)
a
1.1
14.3
−7
rate in feed = 56 × 10 mol/h.
3
Reaction conditions: 1000 psi, 0.03% propyl ether in 250 cm (STP)/
2
min H , 12.5 m catalyst.
2
the propanol conversion reported is the “apparent” conver-
sion, since propanol could also be formed from the reaction
of the aldehyde. However, the propanol formation over the
support is expected to be quite small, making the “appar-
ent” conversion of propanol close to the real conversion. The
conversion of propanal does not appear to have been affected
significantly by the presence of alcohol in the feed. The main
products are 2-methyl-pentanal, which is the result of con-
densation of propanal and propanol, and propyl ether, which
results from the reaction of two propanol molecules. The
other products are propyl-propionate and propionic acid. Ac-
etals and some trimers are also among the products. An
important effect of the presence of alcohol is seen in the de-
creased yield of 2-methyl-pentenal. Reaction between two
aldehyde molecules is somewhat hindered by reaction be-
tween aldehyde and alcohol molecules, as reflected in the
increased products of propyl-propionate, 2-methyl-pentanal,
and propyl ether.
3
.5. Effect of support and catalyst composition on aldehyde
reaction network
3
.5.1. Role of the support
To assess the role of the support in the reaction network,
we performed propanal hydrogenation experiments over a
bare support, which was sulfided by the same procedure that
was used for catalysts. Table 11 clearly shows that a sul-
fided γ -Al2O3 support has significant activity in converting
propanal. The major product is 2-methyl-pentenal, which re-
sults from aldol condensation of propanals. Its selectivity
◦
is over 95% in the temperature range of 140–180 C. The
second highest selectivity is for trimers. The hydrogenation
activity is negligible, with only trace amounts of propanol
produced. Propyl-propionate appears at higher temperatures.
The effect of propanol on the reaction of propanal over
sulfided γ -Al2O3 support is presented in Table 12. The prod-
uct distributions in this table are presented as flow rates.
Unlike the conversion over the catalyst, the propanol conver-
sion is much higher over the support. It should be noted that
3.5.2. Effect of catalyst composition
The effect of metal loading on product distribution in
propanal hydrogenation over sulfided (Ni)Mo/γ -Al2O3 cat-

X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
29
Table 13
Effect of metal loading on product distribution in propanal hydrogenation
a
over sulfided catalysts
Products
% NiO–% MoO /Al O catalysts
3 2 3
0–4
0–10 0–15 3–15 7–15
Ethane
Propane
Propanol
0.1
0.1
32.3 56.3
0.1
0.6
0.1
0.8
67.6
0.1
3.6
6.3
0.2
–
0.5
1.8
76.1
1.5
1.8
2.1
0.2
–
2.2
7.5
75.9
1.8
0.1
0.1
1.4
–
2-Methyl-pentane
2-Methyl-pentene
2-Methyl-2-pentene
–
0.1
2.8
4.5
0.4
0.4
9.9
0.3
0.6
1.6
7.2
0.4
0.2
10.7
1.2
2.3
1.0
Propyl ether
Propionic acid
2
-Methyl-pentanal
Propyl-propionate
-Methyl-2-pentenal
Propionaldehyde dipropyl acetal
6.1
–
–
2.8
–
–
1.8
–
–
2
–
–
–
–
2
-Methyl-pentanol
32.8 21.0
12.4
0.6
2.2
11.8
0.3
1.0
7.7
0.2
1.3
2-Methyl-2-pentenol
2.2
8.0
0.7
2.4
Trimers
Fig. 7. Propanal TPD profiles over sulfided 3% Ni–15% Mo/Al O catalyst
2
3
and sulfided Al O support.
2
3
Conversion of propanal (%)
a
55.3 76.1
82.7
97.2
99.6
◦
3
Reaction conditions: 1000 psi, 180 C, 0.26% propanal 250 cm (STP)/
◦
2
29) desorption feature centered at 117 C (Fig. 7). The ad-
sorption sites are the OH groups on the Al2O3 surface. The
profile obtained from the catalyst shows two major propanal
desorption features at 95 C and 158 C. The peak at the
lower temperature may represent propanal desorbing from
the exposed Al2O3 surface and/or Brønsted acid sites asso-
min, 25 m catalyst.
alysts is presented in Table 13. It can be seen that, with
increasing Mo loading, that hydrogenation activity increases
significantly. It appears that conversion of propanal and se-
lectivity for propanol increase markedly, whereas the selec-
tivity for hydrocarbons, which mainly result from cracking
and full hydrogenation over CUS, shows a more gradual in-
crease. Meanwhile, the selectivity for heavy O-containing
species (e.g., 2-methyl-pentanal, 2-methyl-pentanol), which
are mainly produced from the reactions over the exposed
γ -Al2O3 surface, decreases pronouncedly. With the addition
of Ni, the conversion of propanal and selectivity for propanol
are enhanced further. Selectivity for heavy O-containing
products decreases, whereas the selectivity for light hydro-
carbons increases, especially at higher Ni loading.
◦
◦
◦
ciated with Mo centers. The desorption temperature (95 C),
which is lower for the catalyst than it is for the Al2O3 sup-
◦
port (117 C), may imply that the adsorption of propanal
on Brønsted acid sites is weaker than that on OH groups
over the Al2O3 surface. The broad peak at higher tempera-
◦
tures (around 150 C) indicates the propanal desorbing from
CUS. The broad high-intensity peak of 2-methyl-2-pentenal
(
m/z = 41) observed over the alumina support is consistent
with the reaction network with regard to aldol condensa-
tion of propanal, which is the major reaction of propanal
over the Al2O3 support (Fig. 7). NiMo/Al2O3 catalyst shows
two desorption features for 2-methyl-2-pentenal. The one at
3
.6. Propanal TPD studies on sulfided Al2O3 support and
◦
1
33 C is likely to stem from Brønsted acid sites associated
NiMo/γ -Al2O3 catalyst
with Mo sites. The other one formed at OH groups on the
Al2O3 surface appears at the same temperature as the peak
seen over the Al2O3 support; however, the intensity is much
weaker.
Propanal TPD experiments were carried out over sulfided
% Ni–15% Mo/γ -Al2O3 catalyst and sulfided γ -Al2O3
3
support. Propanal was adsorbed at room temperature, and
the species eluting from the surface as a function of tem-
perature were monitored by GC/MS with the scan mode.
Analysis of individual ions shows that over both the cata-
lyst and the support, the primary species formed during TPD
experiments is 2-methyl-2-pentenal (m/z = 41). 2-Meth-
yl-pentanal/ol (m/z = 43) and propanol (m/z = 31) are
also observed in trace quantities. The desorbed species are
propanal (m/z = 29) and H2O (m/z = 18). H2S (m/z = 34)
desorption is also observed, but only from the catalyst and
not from the support, indicating that Al2O3 cannot be sul-
fided. This result is consistent with our previous observations
with TPD/TPR [17] and XPS [37]. The TPD profile obtained
over the Al2O3 support shows one broad propanal (m/z =
4. Discussion
Reaction experiments using various probe molecules pro-
vide important clues about the reaction network involved in
the hydrogenation of aldehydes over sulfided NiMo/Al2O3
catalysts. It appears that the primary steps involved in the hy-
drogenation of aldehydes are similar regardless of the length
of carbon chain, as seen from the similarity in the trends ob-
served for propanal and hexanal. The reaction networks for
the liquid-phase and gas-phase systems are also similar, as
indicated by the similarity of the product distributions. The

30
X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
Scheme 1. Proposed reaction network of propanal hydrogenation over NiMoS/γ -Al O catalyst.
2
3
primary steps involved in the propanal (A) hydrogenation
network may be envisioned as in Scheme 1. The main reac-
tion is hydrogenation of the C=O double bond to form the
corresponding alcohol (B). Over the sulfided NiMo/Al2O3
catalysts, the primary hydrogenation sites are coordinatively
unsaturated sites (CUS), that is, anionic vacancies associ-
ated with Mo or Ni centers. Our earlier studies on HDN
and HDO catalysis have shown Mo-associated anionic va-
cancies to have substantial hydrogenation activity, but the
intrinsic activity of the Ni-associated CUS sites is much
higher [17,41]. The results from our present investigation
suggest that the conclusions derived earlier about the hydro-
genation function of the NiMoS catalysts hold for aldehyde
hydrogenation reactions as well. The correlations observed
between the NO adsorption capacities of these catalysts and
the alcohol formation rates provide further evidence about
the role of anion vacancies in the transformation of aldehy-
des to alcohols [37–39].
Scheme 2. Proposed mechanism for aldol condensation of propanals over
acid–base sites.
One of the major side reactions is aldol condensation
of two aldehyde molecules. Aldol condensation reactions,
which are used for the production of many fine chemicals,
are primarily base-catalyzed [43, and references therein].
However, acid–base bifunctional catalysts can also promote
aldol condensation reactions [44–47]. The existence of OH
groups with different acidities/basicities was demonstrated
with the DRIFTS technique and is reported in our previ-
ous articles [37,38]. Over alumina support, therefore, aldol
condensation is likely to occur, making use of the bifunc-
tionality of the OH groups (Scheme 2). The weak acid sites
can activate the carbonyl group through polarization of the
C–O bond, which leaves carbon of the aldehyde group with
a higher positive charge density (W). The basic sites, on

X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
31
Scheme 3. Proposed network for hydrogenation of α, β-unsaturated aldehyde.
the other hand, can form enolate ions (X), which can act
as carbon nucleophiles (carbanion) and attack the positively
charged carbon of the intermediate formed on the acidic
sites. The aldol intermediate (C) can then lose water and
form a α, β-unsaturated carbonyl compound (D). As seen
in Table 11, the main species formed over the bare alumina
support is the product of this reaction, namely 2-methyl-2-
pentenal.
further hydrogenate into saturated alcohol (F), as seen in
Table 6. Saturated alcohols, in turn, can lead to saturated
hydrocarbons (H), as seen in Table 7. All of the species rep-
resented in this scheme except the enolic intermediate (U)
were observed in the propanal hydrogenation product stream
over NiMoS/Al2O3 catalysts. These species were the pri-
mary products when 2-methyl-2-pentenal was used as the
feed molecule (Table 8).
Whereas “self-condensation” is an important side reac-
tion in the hydrogenation network of linear aldehydes, re-
actions carried out with branched carbon chains show very
different results. This difference is mainly due to the role of
α-C in linear aldehyde molecules. As seen previously, the
linear aldehydes are much more prone to condensation re-
actions, leading to heavy product formation. In aldehydes
with branched carbon chains, such as 2-ethyl-butanal and
The second important group of side reactions in the
propanal hydrogenation network (Scheme 1) involves those
between aldehyde and alcohol species. These reactions can
also lead to the formation of 2-methyl-pentanal (E) (Tables 5
and 12). An addition reaction between a propanal and a
propanol molecule over an acidic site can lead to a hemi-
acetal intermediate (I). Hemiacetal, which is unstable, gives
a propionaldehyde dipropyl acetal species (J) when reacted
with another alcohol molecule. Propionaldehyde dipropyl
acetal can undergo hydrocracking to give dipropyl ether (K)
and propanol (Table 9). Dipropyl ether can further react to
give propane (L) and propanol, as seen in Table 10.
2
-methyl-pentanal, α-H is not readily accessible, making the
reactions between two aldehyde molecules or an aldehyde
and an alcohol molecule much more difficult. Therefore,
these molecules do not readily participate in reactions that
lead to the formation of heavy products, exhibiting a steric
hindrance effect. Table 6 shows that no heavy products are
produced from the reaction of 2-methyl-pentanal, except for
small amounts of ether from the dehydration of 2-methyl-
pentanol.
α, β-unsaturated aldehyde, which is the major primary
by-product from aldol condensation of propanal, is a very ac-
tive molecule and can further undergo various hydrogenation
reactions [48]. Scheme 3 presents various hydrogen addi-
tion reactions that the 2-methyl-2-pentenal (D) molecule can
undergo over NiMoS/Al2O3 catalysts. The hydrogenation
of the C=O double bond (1, 2 addition) gives the unsat-
urated alcohol (T), which can further hydrogenate into the
saturated alcohol (F). It can also undergo hydrogenolysis to
form the corresponding unsaturated hydrocarbon (G). The
Heavy products observed in the reaction product stream
(M, N, O) are the results of additional condensation reactions
involving C6 aldehydes and alcohols further reacting with C3
aldehydes.
While the evidence seen in our reaction studies and our
TPD experiments clearly point to the role of the bare sup-
port in the catalysis of heavy product formation, it is not
clear whether OH groups alone are responsible for these re-
actions. Studies in the literature have reported the activity
of Al3 sites in catalyzing condensation reactions of ace-
tones [49,50]. It is known that the condensation of aldehydes
can also be catalyzed by liquid [51] and solid [46,52] acid
catalysts. Although the heavy product formation rates cor-
relate very well with the density of surface OH groups, the
contribution of other sites cannot be completely ruled out.
The Brønsted acid sites associated with Mo centers can also
be involved in condensation reactions, although their con-
tribution is likely to be small. In our earlier papers, heavy
product formation rates were seen to correlate with the CO2
uptake capacity of the catalysts [37,38]. However, the fact
+
1
,4-addition of hydrogen gives the enolic intermediate (U),
which quickly isomerizes into saturated aldehyde (E) or fur-
ther reacts with hydrogen to give unsaturated hydrocarbon
molecules (V). The 3,4 addition of hydrogen can also give
the saturated aldehyde (E) directly. Saturated aldehyde can

32
X. Wang et al. / Journal of Catalysis 231 (2005) 20–32
that the heavy product formation rates do not extrapolate to
the origin, even when CO2 uptake is zero, may suggest the
contribution of sites other than the OH groups to these reac-
tions.
[9] H. Topsøe, B.S. Clausen, F.E. Massoth (Eds.), Hydrotreating Cataly-
sis, Springer-Verlag, Berlin, 1996.
10] T.C. Ho, Catal. Rev.-Sci. Eng. 30 (1988) 117.
11] C.N. Satterfield, Heterogeneous Catalysis in Industrial Practice, sec-
ond ed., McGraw-Hill, New York, 1991, p. 383, Chap. 9.
12] M.J. Girgis, B.C. Gates, Ind. Eng. Chem. Res. 30 (1991) 2021.
[
[
Other minor reactions involved in the aldehyde hydro-
genation network (Scheme 1) in the presence of water in-
clude the reaction of propanal with water to give propionic
acid (P). Further reaction with propanol may lead to ester
formation (Q). Liquid-phase data obtained with and without
water provide evidence of these reaction steps (Fig. 6).
Light products such as C1–C2 alkanes can be produced
from the decarbonylation of propanal. Complete hydrogena-
tion followed by cracking can also lead to the formation of
small hydrocarbons. Light products account for only a very
small percentage of the overall yield loss for these catalysts
and show an increase with increasing Ni loading.
[
[13] G. Perot, Catal. Today 10 (1991) 447.
[
[
14] E. Laurent, B. Delmont, J. Catal. 146 (1994) 281.
15] R. Prins, in: G. Ertl, H. Knozinger, J. Weitkamp (Eds.), Hydrode-
sulfurization, Hydrodenitrogenation, Hydrodeoxygenation, and Hy-
drodechlorination, Handbook of Heterogeneous Catalysis, VCH,
Weinheim, 1997.
[16] S.C. Kim, F.E. Massoth, Ind. Eng. Chem. Res. 39 (2000) 1705.
17] L. Zhang, G. Karakas, U.S. Ozkan, J. Catal. 178 (1998) 457.
18] A. Bunch, L. Zhang, G. Karakas, U.S. Ozkan, Appl. Catal. A: Gen. 190
[
[
(2000) 51.
[
[
19] A. Bunch, U.S. Ozkan, J. Catal. 206 (2002) 177.
20] A.M. Escudey-Castro, L. Broussiers McLeeod, E.J. Gil-Llambias,
Appl. Catal. 4 (1982) 371.
[
21] F.E. Massoth, in: Proc. 4th Int. Conf.: The Chemistry and Uses of
Molybdenum, Golden, Colorado, 1982.
[
[
[
[
[
[
[
22] J. Maternova, Appl. Catal. 3 (1982) 3.
23] G. Muralidhar, F.E. Massoth, J. Shabtai, J. Catal. 85 (1984) 44.
24] F.E. Massoth, J. Catal. 36 (1975) 164.
25] E. Payen, S. Kasztelan, J. Grimblot, J. Mol. Struct. 174 (1988) 71.
26] N.-Y. Topsøe, H. Topsøe, J. Catal. 139 (1993) 641.
27] N.-Y. Topsøe, H. Topsøe, F.E. Massoth, J. Catal. 119 (1989) 252.
28] S.H. Yang, C.N. Satterfield, J. Catal. 81 (1983) 168.
5
. Conclusions
Propanal is shown to be a representative model com-
pound for the study of the hydrogenation reaction of linear
aldehydes. The main reaction in the aldehyde hydrogena-
tion process is the hydrogenation of the C=O double bond,
which takes place over the coordinatively unsaturated sites.
The major side reactions are self-condensation of aldehy-
des and condensation of aldehydes with alcohols. Both re-
actions involve α-hydrogen and are primarily catalyzed by
acid–base bifunctional sites over the exposed Al2O3 sur-
faces. In aldehyde molecules with a branched carbon chain,
such as 2-ethyl-butanal and 2-methyl-pentanal, α-hydrogen
is not readily accessible, making the condensation reactions
much more difficult. The steric hindrance effect exhibited
by branched-chain aldehydes is the main reason for the
low heavy-product selectivities. The similarities between the
trends observed in the gas phase and the liquid phase imply
that the experimental results obtained in the gas phase are
still relevant for industrial applications of similar reactions,
which use primarily liquid feeds.
[29] J. Maternova, Appl. Catal. 6 (1982) 61.
[30] V. Stuchly, L. Baranek, Appl. Catal. 35 (1987) 35.
[31] L. Vivier, S. Kasztelan, G. Perot, Bull. Soc. Chim. Belg. 100 (1991)
01.
8
[32] J. Miciukiewicz, W. Zmierczak, F.E. Massoth, Bull. Soc. Chim.
Belg. 96 (1987) 915.
[33] E. Furimsky, Appl. Catal. 6 (1983) 159.
[34] M.S. Rana, B.N. Srinivas, S.K. Maity, G. Murali Dhar, T.S.R. Prasada
Rao, J. Catal. 195 (2000) 31.
[35] M.S. Rana, S.K. Maity, J. Ancheyta, G. Murali Dhar, T.S.R. Prasada
Rao, Appl. Catal. 258 (2004) 215.
[36] M.W. Vogelzang, C.L. Li, G.A. Schuit, B.C. Gates, L. Petrakis,
J. Catal. 84 (1983) 170.
[37] X. Wang, G. Li, U.S. Ozkan, J. Mol. Catal. 217 (2004) 219.
[
[
[
38] X. Wang, U.S. Ozkan, J. Catal. 227 (2004) 492.
39] X. Wang, U.S. Ozkan, J. Mol. Catal., submitted for publication.
40] U.S. Ozkan, S. Ni, L. Zhang, E. Moctezuma, Energy Fuels 8 (1994)
249.
[41] U.S. Ozkan, L. Zhang, S. Ni, E. Moctezuma, J. Catal. 148 (1994) 181.
[42] U.S. Ozkan, Y. Cai, M.W. Kumthekar, L. Zhang, J. Catal. 142 (1993)
182.
[
[
43] B.F. Sels, D.E. De Vos, P.A. Jacobs, Catal. Rev. 43 (2001) 443.
44] M.J. Climent, A. Corma, R. Guil-Lopez, S. Iborra, J. Primo, J. Ca-
tal. 175 (1998) 70.
References
[
1] J.J. MeKetta, W.A. Cunningham (Eds.), Encyclopedia of Chemical
Processing and Design, vol. 33, New York and Basel, 1990, p. 46.
2] P. Grange, Catal. Rev.-Sci. Eng. 21 (1980) 135.
[45] M.J. Climent, A. Corma, R. Garcia, R. Guil-Lopez, S. Iborra, V. For-
nes, J. Catal. 197 (1998) 385.
[46] M.J. Climent, A. Corma, S. Iborra, A. Velty, J. Mol. Catal. A 182–183
(2002) 327.
[
[
3] J. Laine, F. Severino, R. Golding, J. Chem. Technol. Biotechnol. A 34
(
1984) 387.
[47] D. Tichit, D. Lutic, B. Coq, R. Durand, R. Teissier, J. Catal. 219 (2003)
167.
[
4] H. Topsøe, B.S. Clausen, Catal. Rev.-Sci. Eng. 26 (1984) 395.
[
5] H. Topsøe, B.S. Clausen, R. Candia, C. Wivel, S. Morup, J. Catal. 68
(
[48] P. Gallezot, D. Richard, Catal. Rev.-Sci. Eng. 40 (1998) 81.
[49] B.E. Hanson, L.F. Wiesermand, G.W. Wagner, R.A. Kaufman, Lang-
muir 3 (1987) 549.
[50] A.G. Panov, J.J. Fripiat, J. Catal. 178 (1998) 188.
[51] C.R. Hauser, D.S. Breslow, 62 (1940) 2389.
[52] P. Mastagli, G. Lagrange, Compt. Rend. 244 (1957) 207.
1981) 433.
6] C. Wivel, R. Candia, B.S. Clausen, S. Morup, H. Topsøe, J. Catal. 87
1984) 497.
[
(
[
[
7] H. Topsøe, B.S. Clausen, Appl. Catal. 25 (1986) 273.
8] N.-Y. Topsøe, H. Topsøe, J. Catal. 84 (1983) 386.