2
8
M.K. Gnanamani et al. / Journal of Catalysis 277 (2011) 27–35
high-pressure fixed-bed reactor. DRIFTS experiments were per-
formed for the support and catalyst to examine the surface species
generated during the adsorption of reactants at elevated tempera-
tures. By comparing these results, a plausible mechanism was pro-
posed for the hydrogenation of ethyl butyrate and butyric acid over
the total selectivity of all carbon-containing products equaled
100%. The product gas stream exiting the reactor system was
passed through a trap at 273 K to separate liquid and gas products.
1 4
Gas-phase products (C –C ) were analyzed using a micro GC (HP
Quad series, Refinery Gas Analyzer) equipped with a TCD detector,
while the liquid products condensed in the 273 K trap were ana-
lyzed separately using a HP 5890 GC with DB-5 capillary column.
The deuterium switch was made after reaching steady-state con-
version of reactants (EB, BA, and butyraldehyde) in order to mea-
sure the deuterium KIE for hydrogenation of ethyl butyrate,
butyric acid, and butyraldehyde, respectively. The products con-
taining deuterium (especially 1-butanol and ethyl alcohol) were
analyzed by GC–MS.
2
2 3.
5% Co/c-Al O
2
. Experimental
The catalyst used in this study was cobalt supported on
c-Al
2 3
O ,
which was prepared from a commercially available c-Al O sup-
2 3
port (Catalox 150). Cobalt was introduced to the support by the
slurry phase impregnation method, whereby the aqueous solution
of cobalt nitrate was 2.5 times the pore volume of the support. Due
to the solubility limit of cobalt nitrate, two sequential impregna-
tion and drying steps were used, the latter carried out using a
rotary evaporator under vacuum. After impregnation with
3. Results and discussion
3.1. Hydrogenation of ethyl butyrate (EB)
2
5 wt.% of cobalt, the catalyst was dried at 373 K and then calcined
in flowing air at 623 K for 4 h.
While the hydrogenation of ethyl butyrate is expected to give 1-
butanol and ethyl alcohol, measurable amounts of butyl butyrate,
C –C paraffins, and ethoxy butane were also observed in the prod-
1 4
ucts (Scheme 1). Conversion and selectivity data for hydrogenation
of ethyl butyrate are provided in Table 1.
The percentage conversion of ethyl butyrate was high (ꢂ45%)
during the initial period of time and reached a steady value after
about 24 h of time on-stream. The selectivity to ethanol was
42.0–44.0%, while the selectivity to 1-butanol was 30.0–32.0%. In
addition to alcohols, different side reactions (e.g., transesterifica-
tion, dehydration, cracking) produced a measurable amount of bu-
Nitrogen physisorption over the catalyst sample was carried out
at 77 K using a Micromeritics Tri-Star instrument. The sample was
first outgassed overnight at 433 K to 50 mTorr. The specific surface
area of the catalyst was estimated by the BET method to be
2
3
1
03 m /g, while the single point pore volume was 0.258 cm /g
cat. and the average pore radius was 5.0 nm.
DRIFTS experiments were carried out using a Nicolet Nexus 870
equipped with a DTGS–TEC detector to monitor the various surface
species produced during the adsorption of reactants following acti-
2
vation with H at elevated temperatures. A high-pressure/high-
temperature chamber fitted with ZnSe windows was utilized as
the catalyst holder. The gas lines leading to and from the IR cell
were heat traced, insulated with ceramic fiber tape, and further
covered with general purpose insulating wrap. Ethyl butyrate
tyl butyrate, ethoxy butane, and linear C
products.
Deuterium switching was carried out after reaching a desirable
conversion of ester in order to determine the nature of the isotope
effect that exists for hydrogenation of ethyl butyrate. Ethyl buty-
1 4
–C alkanes as by-
(
EB) and n-butyric acid (BA) were introduced to the IR cell using
L injections with a gas-tight syringe, and the carrier gas used
was helium. Scans were taken at a resolution of 4 to give a data
5
l
rate conversion increased by a factor of ꢂ1.81 (i.e., k
H D
/k = 0.55)
and then returned to the baseline conversion level after switching
back to hydrogen. The data in Table 2 and Fig. 1 demonstrate that
hydrogenation of ethyl butyrate displays an inverse isotope effect
and this deuterium isotope effect also exists for the rate of forma-
tion of 1-butanol and ethanol as well. The appearance of an inverse
isotope effect for hydrogenation of ethyl butyrate to alcohol indi-
cates that a hydrogen addition step is involved in the rate-deter-
mining step. In several studies of the hydrogenolysis of methyl
and ethyl acetates, the authors advocate that hydrogenation pro-
ceeds via dissociatively adsorbed acetates [18]. Many suggest that
hydrogenation of the acyl fragment is the rate-determining step
ꢀ1
spacing of 1.928 cm . Typically, 512 scans were taken to improve
the signal to noise ratio. The catalyst was initially activated with H
2
at 773 K for 12 h before adsorbing the probe molecule (EB or BA) at
the desired temperature. Similar experiments were performed over
c
-Al
behavior of EB between the support (
Co/ -Al ). H -TPD was carried out after adsorbing EB on
-Al and 25% Co/ -Al at 473 K to determine the stability
2
O
3
as well in order to see the differences in the adsorbing
c
-Al ) and catalyst (25%
2 3
O
c
2
O
3
2
c
2
O
3
c
2 3
O
of the intermediate species formed over the surface of the catalyst
in the temperature range of 473–623 K.
Reactions were conducted using a microcatalytic reactor system
capable of operating at high pressures and temperatures. Liquid
feed at the desired space velocity was introduced using a Milton
Roy mini-piston pump. Hydrogen gas flow rates were metered
using mass flow controllers (Brooks Instruments Model 5810B).
The reactor portion of this system consisted of a 1/2 ꢁ 21 inch
plug-flow microreactor. The catalyst was held in place by a bed
of quartz wool. Typical catalyst charges in this unit were approxi-
during acetate hydrogenolysis over Cu/SiO
2
catalysts [9,11].
In the case of 25% Co/ -Al , the first question is whether the
c
2 3
O
ethyl butyrate is adsorbed associatively or dissociatively. It is also
important to shed light on the roles played by both the metal (Co)
and support (c-Al O ) in adsorbing and turning over reactants and
2 3
intermediates on the catalyst surface. Thus, ethyl butyrate adsorp-
tion was investigated on both the unpromoted support and cobalt
catalyst by DRIFTS. These findings were then compared to the cat-
alytic testing results to gain insight into the likely rate-determining
step.
mately 1.5 g of unreduced 25% Co/
diluent. All chemicals (i.e., ethyl butyrate (EB), n-butyric acid
BA), and n-butyraldehyde) were of high purity (>99.5%) and pur-
chased from Sigma Aldrich. The catalyst was activated for 15 h at
23 K using a H :He (1:3) mixture. Total conversion of reactants
2 3
c-Al O along with 3.0 g of SiC
(
DRIFTS spectra for ethyl butyrate adsorption over
2 3
c-Al O and
25% Co/ -Al at various conditions are displayed in Fig. 2. Table 3
c
2 3
O
6
2
summarizes the vibrational bands and mode assignments for sur-
face species formed from adsorption of ethyl butyrate followed
for each experiment was kept below 15–20% in order to keep the
reaction under differential operating conditions except during
the hydrogenation of n-butyraldehyde.
Conversion has been defined as number of moles of ester con-
verted per mole of ester introduced. To account for the change in
number of moles during reaction, selectivity was defined such that
by the introduction of H
Al . At 473 K under He,
3000 cm region corresponding to CH
[e.g., (CH)] from adsorbed ethoxy and butanoate species. Charac-
teristic (C–O) bands corresponding to mono-
2
at 493 K over
-Al exhibits bands in the 2760–
and CH vibrational modes
2 3
c-Al O and 25% Co/c-
2
O
3
c
2 3
O
ꢀ1
3
2
m
m
(m