Ruthenium-supported catalysts for the stereoselective hydrogenation of paracetamol to 4-trans-acetamidocyclohexanol: Effect of support, metal precursor, and solvent

Article abstract of DOI:10.1016/j.jcat.2004.11.018

The influence of the support, the metal precursor, and the solvent on the selective hydrogenation of paracetamol (4-acetamidophenol) was studied over supported ruthenium catalysts. The catalysts supported on the oxidic supports Al₂O₃ and SiO₂ gave the best results in terms of activity, selectivity for the acetamidocyclohexanols (99%), and stereoselectivity for the trans isomer (53 and 46%, respectively). Carbon-supported catalysts produced larger amounts of secondary compounds, mainly N-cyclohexylacetamide, which was derived from the hydrogenolysis reaction of the OH group. The use of a chloride precursor resulted in the enhancement of the formation of N-cyclohexylacetamide and partially hydrogenated products; the stereoselectivity also increased. Moreover, because of the acidity caused by residual Cl, condensation led to oligomers of paracetamol. In spite of the decrease in the selectivity for cyclohexanol derivatives when the more polar solvent ethanol was used instead of isopropanol or tetrahydrofuran the stereoselectivity for the trans isomer increased from 30 to 38%. The results confirm that the factors studied affect the mode of adsorption of the molecule of paracetamol on the catalyst in different ways. These effects determine the product distribution and the selectivity of the reaction.

Full text of DOI:10.1016/j.jcat.2004.11.018

Journal of Catalysis 229 (2005) 439–445
www.elsevier.com/locate/jcat
Ruthenium-supported catalysts for the stereoselective hydrogenation
of paracetamol to 4-trans-acetamidocyclohexanol:
effect of support, metal precursor, and solvent
B. Bachiller-Baeza a,∗, A. Guerrero-Ruíz b, I. Rodríguez-Ramos a
a
Instituto de Catálisis y Petroleoquímica, CSIC, c/ Marie Curie 2, Campus de Cantoblanco, 28049 Madrid, Spain
b
Departamento de Química Inorgánica y Química Técnica, Facultad de Ciencias, UNED, c/ Senda del Rey, no 9, 28040 Madrid, Spain
Received 14 September 2004; revised 12 November 2004; accepted 15 November 2004
Available online 23 December 2004
Abstract
The influence of the support, the metal precursor, and the solvent on the selective hydrogenation of paracetamol (4-acetamidophenol)
was studied over supported ruthenium catalysts. The catalysts supported on the oxidic supports Al O and SiO gave the best results
2
3
2
in terms of activity, selectivity for the acetamidocyclohexanols (99%), and stereoselectivity for the trans isomer (53 and 46%, respec-
tively). Carbon-supported catalysts produced larger amounts of secondary compounds, mainly N-cyclohexylacetamide, which was derived
from the hydrogenolysis reaction of the OH group. The use of a chloride precursor resulted in the enhancement of the formation of N-
cyclohexylacetamide and partially hydrogenated products; the stereoselectivity also increased. Moreover, because of the acidity caused by
residual Cl, condensation led to oligomers of paracetamol. In spite of the decrease in the selectivity for cyclohexanol derivatives when the
more polar solvent ethanol was used instead of isopropanol or tetrahydrofuran the stereoselectivity for the trans isomer increased from 30
to 38%.
The results confirm that the factors studied affect the mode of adsorption of the molecule of paracetamol on the catalyst in different ways.
These effects determine the product distribution and the selectivity of the reaction.
2004 Elsevier Inc. All rights reserved.
Keywords: Stereoselective hydrogenation; Cis and trans isomers; 4-Aminocyclohexanol; Ruthenium-supported catalysts
1. Introduction
in the hydrogenation reaction and its stereoselectivity [1–8].
Group VIII metals are generally used, but different behaviors
are observed. Whereas Pd gives mainly the cyclohexanone
derivatives, rhodium and platinum are very selective for the
cis isomer. Nickel is more stereoselective for the trans iso-
mer but generates higher amounts of hydrogenolytic cleav-
age products. However, exhaustive studies of the range of
parameters that can influence the reaction (support, reaction
conditions, promoters, or additives), as have been carried
out for the hydrogenation of benzene or phenol [9–11], are
scarce.
Among the substituted cyclohexanols, the synthesis of
4-trans-aminocyclohexanol is of great importance. This
compound is an intermediate in the multistep synthesis of
Ambroxol [12–14], N-(2-amino-3,5-dibromobenzyl)-trans-
4-aminocyclohexanol, which is a bronchosecretolitic agent
The hydrogenation of disubstituted aromatics, particu-
larly of substituted phenols to the corresponding substituted
cyclohexanones and cyclohexanols, has attracted the inter-
est of the industry and researchers. The reason for this is that
these compounds are used as key intermediates for the man-
ufacture of fine chemicals for the fragrance and detergent
industries. Moreover, the stereochemistry of the reactions
is important since generally only one of the isomers of the
cis and trans mixture has the desired properties. Several re-
ports have dealt with the role of the nature of the metal phase
*
Corresponding author. Fax: +3491 585 4760.
E-mail address: b.bachiller@icp.csic.es (B. Bachiller-Baeza).
0021-9517/$ – see front matter
doi:10.1016/j.jcat.2004.11.018
2004 Elsevier Inc. All rights reserved.

440
B. Bachiller-Baeza et al. / Journal of Catalysis 229 (2005) 439–445
for the prevention and treatment of diseases of the res-
piratory tract and bronchial tubes [15,16]. Although sev-
eral procedures have been tested for obtaining trans-4-
aminocyclohexanol (i.e., hydrogenation of p-nitrophenol or
p-aminophenol), the main route is the catalytic hydrogena-
tion of 4-acetamidophenol (paracetamol) [17–19] to yield
a cis/trans mixture of 4-acetamidocyclohexanol, which is
subsequently hydrolyzed to cis/trans-4-aminocyclohexanol;
the isomers are finally separated by fractional crystalliza-
tion. Considering the complete process and that the product
is used in pharmacology, the aim is to achieve the maximum
yields of the trans isomer in the initial hydrogenation reac-
tion to reduce the amounts of products of no interest. Apart
from the economic and environmental benefits, this would
simplify the workup and purification of the reaction product.
The purpose of this study was to investigate the factors
that can affect the hydrogenation of paracetamol in the liquid
phase over ruthenium catalysts. The effects of the support
(metal oxides or carbon materials) and the ruthenium metal
precursor and the role of the reaction solvent were investi-
gated.
chromatography (Varian 3400) with a thermal conductivity
detector (TCD).
The metal dispersion of the catalysts was determined
from H2 chemisorption measurements at 298 K in a volu-
metric system, described elsewhere [21]. The samples were
subjected to in situ reduction pretreatment in H2 at 673 K.
Mean particle sizes were determined under the assumptions
of M:H2 = 2:1 stoichiometry and spherical metal particles
¯
(d (nm) = 1.32/D) [22].
The heat of CO adsorption at 330 K was measured in
a differential heat-flow microcalorimeter (Tian–Calvet type
C80 from Setaram). The adsorption vessels were connected
to a Pyrex volumetric apparatus equipped with greaseless
stopcocks that permitted the introduction of small pulses
of CO [23]. The samples were activated in flowing H2 at
673 K for 2 h and, after outgassing at this temperature for
16 h, were cooled to the adsorption temperature. Succes-
sive doses of CO were then transferred to the sample. The
equilibrium pressure was measured by means of a Baratron
pressure transducer MKS instrument. The calorimetric and
volumetric data were stored and analyzed with a microcom-
puter.
2.3. Catalytic hydrogenation
2. Experimental and methods
The hydrogenation reaction was carried out under high
pressure in an autoclave fitted with a magnetically coupled
stirring head, inlet/outlet ports, and a pressure gauge. Af-
ter being reduced in flowing hydrogen at 673 for 2 h, the
catalyst (500 mg) was transferred to the reactor flask, which
contained a solution of paracetamol (98% 4-acetamidofenol;
Research Chemical, Ltd.) in ethanol (ca. 25 µmolml−1).
The autoclave was closed, purged in He, and heated to
the reaction temperature (393 K). Then the autoclave was
pressurized to 50 bar with H2 and the reaction started.
Samples were taken periodically and analyzed with a gas
chromatograph (Varian 3800) equipped with a flame ion-
ization detector (FID) and a capillary column (MFE-624,
25 m × 0.25 mm × 0.25 µm; initial temperature, 155 ◦C,
5 min; final temperature 200 ◦C, 6 min; rate 10 ◦C min−1).
Preliminary runs carried out at different rates of stirring, lev-
els of loading, and catalyst grain size revealed the absence of
external and internal mass transfer limitations.
2.1. Catalyst preparation
The supports were silica (Aerosil Degussa, SBET = 180
m2 g−1); alumina (Puralox Condea, SBET = 192 m2 g−1);
a molecular sieve of carbon, S (SBET = 970 m2 g−1); and an
activated carbon (ICASA, Spain, SBET = 961 m2 g−1). The
S carbon was synthesized by carbonization of the copolymer
Saran (90% vinyliden chloride–10% vinyl chloride; Aldrich)
under inert conditions (N2; 100 cc min−1) at 1273 K [20].
The Ru-supported catalysts (2 wt% metal loading) on
alumina (RuA), silica (RuSi), and activated carbon (RuC)
were prepared by the incipient wetness technique with aque-
ous Ru(NO)(NO3)3 solutions. Another catalyst, RuClC, was
prepared with an RuCl3 · xH2O ethanolic solution on acti-
vated carbon. The catalyst supported on S (RuS) was pre-
pared by impregnation of the support (outgassed at 423 K
for 1 h) with a methanolic solution of the Ru(NO)(NO3)3
precursor and subjected to a freeze-thaw cycle. After im-
pregnation, all of the catalysts were dried overnight at 383 K.
Furthermore, the catalysts supported on oxides were cal-
cined in air at 723 K for 3 h.
Considering the possible reaction pathways (Scheme 1),
we calculated the selectivities for alcohols (2 and 2ꢀ) (SH),
N-cyclohexylacetamide (3) (SN), and other products (4)
(SO). The stereoselectivity between the two alcohol isomers
(2 and 2ꢀ) was defined as ST = Ctrans × 100/(Ccis + Ctrans).
2.2. Catalyst characterization
Temperature-programmed reduction (TPR) measure-
ments were carried out in a quartz microreactor over 200
to 300 mg of prepared sample under a continuous flow at
25 ccmin−1 of a H2/He gas mixture (10% H2). The tem-
perature was increased from room temperature to 873 K at
8 K min−1. H2 consumption was measured by on-line gas
3. Results and discussion
3.1. Catalyst characterization
Fig. 1 displays the TPR profiles of H2 consumption for
the Ru-supported catalysts. RuA and RuSi catalysts showed

B. Bachiller-Baeza et al. / Journal of Catalysis 229 (2005) 439–445
441
ꢀ
Scheme 1. Reaction pathways in the hydrogenation of paracetamol (1: paracetamol, 2: trans-4-acetamidocyclohexanol, 2 : cis-4-acetamidocyclohexanol,
3: N-cyclohexylacetamide, 4: 4-acetamidocyclohexanone, and 5: 4-acetamidocyclohexenol).
unique peaks at 405 and 440 K, respectively. The temper-
ature of reduction corresponds to the reduction of RuO2
formed after the calcination of the catalysts; unsupported
bulk RuO2 is reduced in a single peak at 490 K [24]. On the
other hand, the RuC and RuClC catalysts showed several re-
gions of hydrogen consumption: below 423 K, between 423
and 573 K, and above 623 K. The first and second regions
are ascribed to the decomposition of the residual metal pre-
cursor salt, nitrate or chloride, and the reduction of Ru(III)
to Ru0. Reduction of unsupported RuCl3 occurs at tempera-
tures around 433 K [24]. The maxima of the H2 consumption
were observed at 510 K for RuC and at 440 and 508 K for
RuClC. The detection of two peaks for the latter catalyst in-
dicates that two different species are present, one of which is
probably ruthenium oxidized by atmospheric air exposure.
The band in the third region was found only for the RuC
catalysts. Evolution of CH4 and CO, parallel to H2 consump-
tion, indicates partial gasification of the support around the
metal particles [23–26] and decomposition and hydrogena-
tion of the surface oxygen groups [27–30]. These oxygen
groups are generated on the carbon surface during impreg-
nation with the oxidizing solution of the metal precursor
RuNO(NO3)3. The RuS sample showed a peak centered at
Fig. 1. Temperature-programmed reduction profiles for catalysts (a) RuA,
(b) RuSi, (c) RuS, (d) RuC, and (e) RuClC.
540 K (Fig. 1c), indicating that the reduction of the metal
is hindered compared with that of the RuC catalyst. The ab-

442
B. Bachiller-Baeza et al. / Journal of Catalysis 229 (2005) 439–445
Table 1
Catalytic results of ruthenium-supported catalysts on the hydrogenation of
paracetamol at 393 K
a
Catalyst
d
H
Activity
TOF Selectivity (%) Stereoselec-
−1
−1
−1
(s
)
(nm)
tivity, S (%)
)
T
(µmol g
s
S
S
S
O
H
N
M
RuA
RuSi
RuS
RuC
19
33
26
183
56
10
0.26 99
0.14 99
0.02 94
0.01 80
0.06 89
0.02 92
0.01 64
1
1
4
17
10
6
0
0
2
3
1
2
53
46
46
38
33
30
b
2.6
45
d
325
e
91
c
RuClC 29
a
7
24
13 42
Selectivity to cis/trans-acetamidocyclohexanol (S ), N-cyclohexyl-
H
acetamide (S ), other products (S ), and stereoselectivity to trans-aceta-
N
O
midocyclohexanol (S ).
T
b
2–5 nm (determined by XRD and TEM).
Reduction at 723 K and 7.8 nm determined by CO chemisorption.
c
d
e
Solvent IPA.
Solvent THF.
Fig. 2. Evolution of stereoselectivity to trans, S , (---), and of composi-
tion with time in reaction, (2) paracetamol, (1) cis/trans-acetamidocyclo-
hexanol, (!) N-cyclohexylacetamide, (P) other products. Catalyst, RuC;
T
sence of a band at higher temperatures indicates a decrease
in surface gasification and surface oxidation, probably as a
consequence of the differences in the preparation method, as
indicated in Section 2. The higher temperature of reduction
for the two carbon-supported catalysts prepared with the ni-
trate precursor, RuC and RuS, compared with the catalysts
supported on oxides may be due to the presence of oxide-like
Ru species, formed after drying in air and during the precur-
sor decomposition, which interact with the oxidized surface
and consequently are more difficult to reduce.
The results of the H2 chemisorption experiments, carried
out on the ruthenium catalysts, are shown in Table 1. The
catalysts supported on the oxides, RuA and RuSi, showed
low dispersion, probably because of calcination, which fa-
vors the sintering and migration of the RuO2 species. The
chemisorption measurements indicated that 26-nm particles
are present on the RuS catalyst. On the other hand, ruthe-
nium is homogeneously dispersed on the RuC catalyst, with
particles ranging from 2 to 3 nm. The particle size of RuClC
increased to 29 nm. This seems to be a consequence of the
metal precursor used for catalyst preparation. The presence
of some poisoning chloride species may explain the low hy-
drogen adsorption capacity of this sample. Moreover, the
discrepancy between the particle sizes as determined by H2
and CO chemisorption on this catalyst (7.8 nm) [31] support
this assumption.
temperature, 393 K; P (H ) = 50 bar; solvent, ethanol.
2
the hydrogenation reaction of paracetamol can follow the
reaction pathways shown in Scheme 1. The main prod-
ucts obtained under the experimental conditions are a mix-
ture of cis- and trans-4-acetamidocyclohexanol, N-cyclo-
hexylacetamide, and partial hydrogenation products (such as
4-acetamidocyclohexanone). The typical evolution of the re-
action composition with reaction time is displayed in Fig. 2.
Table 1 gives the catalytic results of the hydrogenation of
paracetamol obtained with the Ru catalysts after reduction
at 673 K. The TOF values were calculated from the disper-
sion determined by H2 chemisorption. The data show that
the initial catalytic activity (µmol g−Ru1 s−1) follows the trend
RuA > RuSi > RuC > RuS (Table 1). Moreover, if we take
into account that the metal particle size for the RuS catalyst,
as determined by H2 chemisorption, is overestimated [32],
the general trend indicates that the activity per surface metal
site increases with the metal size. These results agree well
with previous findings reported for Ru catalysts supported
on graphite [35].
As far as the selectivity is concerned, the catalysts sup-
ported on alumina and silica resulted mainly in hydro-
genated products, with selectivities higher than 99% for the
cis and trans alcohols. For the catalysts supported on carbon
materials, the selectivity for the hydrogenated products de-
creased in favor of N-cyclohexylacetamide and products due
to partial hydrogenation, which appeared mostly at the be-
ginning of the reaction. The highest stereoselectivity for the
trans isomer was obtained for the RuA catalyst, with values
around 53%. Whereas high values were also obtained with
the RuSi and the RuS catalysts, a stereoselectivity of only
38% was reached with the RuC sample.
The particle sizes obtained with H2 chemisorption are
consistent with that determined by XRD and TEM for all
of the samples except the RuS catalyst [32]. Both techniques
gave a smaller mean particle size (2–5 nm) for this sample.
3.2. Catalyst reactivity in the hydrogenation of
paracetamol
3.2.1. Effect of support
According to the mechanisms proposed for the hydro-
genation of phenol and alkyl-substituted phenols [3,33,34],
In general, it is postulated that the bonding of an aromatic
to the metal occurs via π-bonds involving electron transfer

B. Bachiller-Baeza et al. / Journal of Catalysis 229 (2005) 439–445
443
from the ring to the unoccupied d metal orbitals [36]. Parac-
etamol is a molecule with a flat structure and has weak acidic
properties; the presence of OH and acetamide substituents in
the aromatic ring has positive resonance effects and would
enhance the adsorption of the aromatic ring by stabilizing
the π-complex. On the catalysts supported on carbon, this
adsorption through the aromatic ring would be stronger be-
cause of the smaller size of the ruthenium particles, leading
to lower activity (Table 1) and to reaction beyond the com-
plete hydrogenation of the ring to produce the cyclohexyl
derivative. However, in addition to aromatic ring interac-
tions, the paracetamol molecule can adsorb via the hydroxyl
group. Assuming a scenario similar to that postulated for
phenol and alkyl-phenols [33,37], the results obtained may
be a consequence of the existence of two adsorbed species
on the catalyst surface: with the ring orientated in a coplanar
or a nonplanar fashion with respect to the surface. On the
RuA and RuSi catalysts, the presence of incompletely re-
duced Ru species (Ruδ+), stabilized by interaction with the
support, should favor the coplanar adsorption of the mole-
cule, leading to complete hydrogenation. Simultaneously,
4-acetamidocyclohexanol may be formed after the stepwise
addition of hydrogen to species adsorbed in the nonplanar
mode. However, the influence of the acid–base properties of
the support on the product distribution, as postulated for the
hydrogenation of phenol and alkyl-phenols, cannot be ruled
out.
Fig. 3. Differential heat of CO adsorption at 330 K for (2) RuC and
(1) RuClC.
The production of trans isomers is generally explained
by isomerization processes or double bond migration of
the partially hydrogenated cyclic alcohol and is associated
with a rollover mechanism [36]. For alkyl-substituted phe-
nols the stepwise addition leads to the alkylcyclohex-1-en-
1-ol, which is not isolated and which is transformed to the
alkylcyclohexanone by tautomerism. Whereas the hydro-
genation of the alkylcyclohexanone leads to both cis and
trans alkylcyclohexanols, the enol is hydrogenated to the
cis isomer, increasing its concentration in the final mix-
ture [3]. Thus, the selectivities for cis and trans isomers,
obtained in the hydrogenation of paracetamol, may indi-
cate that the catalyst influences the keto-enol equilibrium.
On the RuC catalyst, this equilibrium may shift to the enol
form, which is hydrogenated to the cis isomer, increasing its
proportion in the mixture. On the other hand, the absence
of 4-acetamidocyclohexanone shows that the intermediate
is not desorbed and is further hydrogenated to the cyclo-
hexanol derivatives, contrary to what is observed for Pd
catalysts, where the main products are the cyclohexanone
derivatives [3].
However, as indicated above, the mean particle size deter-
mined for RuClC by H2 chemisorption is overestimated.
Moreover, a strong deactivation with time was detected for
the ex-chloride catalysts. Although a similar selectivity for
N-cyclohexylacetamide was obtained, the amount of hydro-
genated products decreased to 64% in favor of an increase in
products of partial hydrogenation and side reactions, proba-
bly hydrogenation of the acetamide group and polymeriza-
tion reactions. In an acidic solution, carbon oligomers of
paracetamol are created by condensation according to the
“head to tail” mechanism [38]. The observed stereoselec-
tivities of the two catalysts differed slightly: 38% for the
RuNO(NO3)3 and 42% for the RuCl3 precursor.
The nature of the metal precursor can have an impact
on the metal surface structure, affecting dispersion, particle
shape, site blocking, or reconstruction [39,40]. In the present
case, slightly lower metal particle sizes are obtained for the
Cl-free catalysts, which agrees well with previous findings.
It has been suggested that this is due to the speciation of
the aqueous solutions used, as RuCl3 solutions contain poly-
meric species, but it may also be due to sintering favored in
the presence of chloride ions [41]. Moreover, the presence of
residual chlorine on the catalyst after the calcination and/or
reduction steps could favor electronic interactions between
Cl and the metal particles or adsorbed species [31,42]. To
study the possible influence of the electronic density of the
metal atoms on the catalyst reactivity, microcalorimetric CO
adsorption experiments were carried out. The profiles of dif-
ferential heat of adsorption (Fig. 3) showed a heterogeneous
3.2.2. Effect of metal precursor
The effect of the metal precursor on the performance of
the catalysts was studied on the carbon-supported sample.
The initial activity (Table 1) for RuClC was lower than that
for RuC, but the TOF value is of the same order, which
does not correlate well with the general trend observed,
that is, an increase in activity with the metal particle size.

444
B. Bachiller-Baeza et al. / Journal of Catalysis 229 (2005) 439–445
distribution of active sites for both catalysts. Although the
heat of adsorption is similar for coverage higher than 0.3, the
catalyst prepared from the nitrate precursor initially gave a
higher initial heat of adsorption. This decrease in the heat of
CO adsorption is explained by electron withdrawal induced
by the chlorine anions in the support–metal interface, result-
ing in a decrease in the back-donation of electrons from Ru
to the π∗ anti-bonding orbital. Shifts to higher wave num-
bers in the IR bands for CO adsorption have been ascribed
to similar electronic modifications related to the presence of
Cl and other electron-withdrawingions [40,43,44]. These in-
teractions seem to result in an enhancement of the adsorption
strength of paracetamol and a decrease in the hydrogenation
rate. A similar negative effect of chlorine has been reported
in the hydrogenation of phenol and styrene, where the hy-
drogenation rates were lower over the palladium catalysts
prepared from chloride salts [45].
Fig. 4. Selectivity to cyclohexanol derivatives (2) and stereoselectivity to
trans isomer (!) vs the dielectric constant of the solvents. Catalyst, RuC;
The acidity of the catalysts, due to the chloride ion, seems
to be responsible for the slight promotion of the hydrogenol-
ysis process and for the deactivation rate; the latter is prob-
ably due to deposition of carbonaceous species on surface
acid sites [36,45]. As suggested above, the observed increase
in selectivity for the trans isomer with the RuClC catalysts
is probably attributable to the modification of the tautomeric
keto-enol equilibrium toward the ketone intermediate as a
result of the interactions between the metal particles and the
residual Cl ions. Modifications of the electronic properties
of ruthenium on alumina-supported catalysts with high chlo-
rine contents seem to account for the decrease in activity and
the higher selectivities for cyclohexene during benzene hy-
drogenation [42].
temperature, 393 K; P (H ) = 50 bar.
2
which may explain the lower activity with EtOH. In this
sense, the diffusivity of paracetamol is known to increase
when surface acidic sites (oxidized carbons, hydrophilic or
polar surfaces, and metal cations) are effectively blocked by
water [38]. On the other hand, the selectivity for the total
hydrogenation of the aromatic ring seems to decrease with
solvent polarity. Moreover, the increase in selectivity for N-
cyclohexylacetamide when EtOH was used could imply a
greater contribution of the activation of the molecule via the
substituent oxygen. This is supported by the favored forma-
tion of phenoxide in polar solvents.
Finally, the results for the RuC catalyst indicate that the
lower the polarity of the solvent is, the higher is the stereo-
selectivity for the cis isomer. This can again be explained by
the favored adsorption of the molecule in a planar mode as
the polarity decreases or by modification of the tautomeric
equilibrium keto-enol of the reaction intermediate.
3.2.3. Effect of the solvent
The catalyst supported on carbon was also selected to
study the influence of the nature of the solvent on catalytic
performance. Two other solvents were tested, 2-propanol
(IPA) and tetrahydrofuran (THF). The initial activity in-
creased in the order EtOH < THF < IPA (Table 1), which
indicates an effect of the solvent. Moreover, as the produc-
tion of the alcohols increased with the two new solvents,
the formation of N-cyclohexylacetamide decreased. On the
other hand, the stereoselectivity for the trans isomer was
lower for IPA and THF than for EtOH.
Effects of solvents have been reported for different het-
erogeneous catalytic processes [7,46,47]. The differences in
activity and selectivity cannot be explained initially by the
values of H2 solubility, since they are very similar for the
three solvents. Thus, we tried to correlate the catalytic results
with the polarity of the reaction medium. Fig. 4 shows the
dependence of the selectivity for acetamidocyclohexanols
and the stereoselectivity for the trans isomer on the dielectric
constant of the solvent. The activity values (Table 1) do not
show a clear relationship with the dielectric constants of the
solvent. However, since the carbon has acidic and basic oxy-
gen surface groups, the more polar solvent may be adsorbed
to the surface, thus reducing the adsorption of the reactant,
4. Conclusions
We have shown that the stereoselective hydrogenation of
paracetamol for 4-trans-acetamidocyclohexanol depends on
several factors. The best results for activity and selectivity
for the cyclohexanols and for the trans isomer were obtained
with the ruthenium catalysts supported on metal oxides. The
influence of the metal precursor on the carbon-supportedcat-
alyst indicates that the residual chloride slightly promotes
the stereoselectivity for the trans isomer. However, it is detri-
mental to the chemo-selectivity, since larger amounts of side
products are obtained, and the catalyst deactivates with time
in reaction. On the other hand, the use of solvents of medium
or low polarity enhances the activity but reduces the stere-
oselectivity. Our results suggested that the product distrib-
ution obtained in the hydrogenation of paracetamol can be
controlled by selecting the appropriate catalyst and reaction
conditions.

B. Bachiller-Baeza et al. / Journal of Catalysis 229 (2005) 439–445
445
Acknowledgments
[21] B. Bachiller-Baeza, I. Rodríguez-Ramos, A. Guerrero-Ruíz,
Langmuir 14 (1998) 3556.
[22] J.R. Anderson, in: Structure of Metallic Catalysts, Academic Press,
San Diego, SA, 1975, p. 295.
[23] A. Guerrero-Ruíz, P. Badenes, I. Rodríguez-Ramos, Appl. Catal. A:
Gen. 173 (1998) 313.
[24] P.G.J. Koopman, A.P.G. Kieboom, H. van Bekkum, React. Kinet.
Catal. Lett. 8 (1978) 389.
The financial support of the MC&T of Spain (project
MAT 2002-04189-C02-01 and 02) is gratefully acknowl-
edged. The authors appreciate valuable discussions with
Dr. J. Martín Medina of Derivados Químicos S.A.
[25] B. Bachiller-Baeza, A. Guerrero-Ruíz, P. Wang, I. Rodríguez-Ramos,
J. Catal. 204 (2001) 450.
[26] B. Bachiller-Baeza, I. Rodríguez-Ramos, A. Guerrero-Ruíz, Appl.
Catal. A: Gen. 205 (2001) 227.
[27] J.L. Figueiredo, M. Pereira, M. Freitas, J. Orfao, Carbon 37 (1999)
1379.
References
[1] G.D. Yadav, P.K. Goel, J. Mol. Catal. A: Gen. 184 (2002) 281.
[2] D.Y. Murzin, A.I. Allachverdiev, N.V. Kul’kova, Stud. Surf. Sci. Catal.
(1993) 243.
[3] J. Tobicík, L. Cerveny, J. Mol. Catal. A: Chem. 194 (2003) 249.
[4] R. Burmeiser, A. Freund, P. Panster, T. Tacke, S. Wieland, Stud. Surf.
Sci. Catal. 92 (1995) 343.
[5] M. Bartók, Stereochemistry of Heterogeneous Metal Catalysis, Wiley,
Chichester, 1985.
[28] J.L. Lemaitre, in: F. Delanay, et al. (Eds.), Characterisation of Hetero-
geneous Catalysts, vol. 28, Marcel Dekker, New York, 1984.
[29] A. Aksoylu, M. Madalena, A. Freitas, M. Pereira, J.L. Figueiredo, Car-
bon 39 (2001) 175.
[30] U. Zielke, K.J. Huttinger, W. Hoffman, Carbon 34 (1996) 983.
[31] B. Bachiller-Baeza, A. Guerrero-Ruiz, I. Rodríguez-Ramos, Appl.
Catal. A: Gen. 192 (2000) 289.
[6] G.V. Smith, F. Notheisz, Heterogeneous Catalysis in Organic Chem-
istry, Academic Press, New York, 1999.
[7] S. Nishimura, Handbook of Heterogeneous Catalytic Hydrogenation
for Organic Synthesis, Wiley, New York, 2001.
[32] B. Bachiller-Baeza, I. Rodríguez-Ramos, A. Guerrero-Ruiz, M. Bes-
son, C. Pinel, J. Mol. Catal. A: Chem. 164 (2000) 147.
[33] M. Bartók, J. Czombos, Stereochemistry of Heterogeneous Metal
Catalysis, Wiley, Chichester, 1985, p. 251.
[34] G. Neri, A.M. Visco, A. Donato, C. Milone, M. Malentacchi, G. Gu-
bitosa, Appl. Catal. A: Gen. 110 (1994) 49.
[8] R.A. Sheldon, in: H. van Bekkun (Ed.), Fine Chemicals through Het-
erogeneous Catalysis, Wiley–VCH, Weinheim, 2001.
[9] C. Park, M.A. Keane, J. Coll. Interface Sci. 266 (2003) 183.
[10] S. Velu, M.P. Kapoor, S. Inagaki, K. Suzuki, Appl. Catal. A: Gen. 245
(2003) 317.
[11] J.W. da-Silva, A.J.G. Cobo, Appl. Catal. A: Gen. 252 (2003) 9.
[12] A. Moroni, Magis Farmaceutici S. R. L., Brescia, US Pat. 4528393,
Italy (1985).
[35] M. Cerro-Alarcon, A. Guerrero-Ruiz, I. Rodríguez-Ramos, Catal. To-
day 93–95 (2004) 395.
[36] A. Kalantar Neyestanaki, P. Mäki-Arvela, H. Backman, H. Karhu,
T. Salmi, J. Väyrynen, D.Yu. Murzin, J. Catal. 218 (2003) 267.
[37] E.-J. Shin, M.A. Keane, J. Catal. 173 (1998) 450.
[38] A.P. Terkyk, G. Rychlicki, S. Biniak, J.P. Lukaszewicz, J. Colloid In-
terface Sci. 257 (2003) 13.
[13] I. Ratz, P. Benko, D. Bozsing, A. Tungler, T. Mathe, G. Kovanyi,
J. Petro, I. Sztruhar, G. Vereczkey, Brit. UK Pat. Appl. GB 2239242,
EGIS Gyogyszergyar, Hungary (1991).
[14] I. Ratz, P. Benko, D. Bozsing, A. Tungler, T. Mathe, G. Kovanyi,
J. Petro, I. Sztruhar, G. Vereczkey, Brit. UK Pat. Appl. GB 2239241,
EGIS Gyogyszergyar, Hungary (1991).
[15] T. Weiser, N. Wilson, Mol. Pharmac. 62 (2002) 433.
[16] B. Yang, D.F. Yao, M. Ohuchi, M. Ide, M. Yano, Y. Okumura, H. Kido,
Eur. Resp. J.: Off. J. Eur. Soc. Clinical Resp. Physiology 19 (2002)
952.
[39] T. Narita, H. Miura, M. Ohira, H. Hondou, K. Sugiyama, T. Matsuda,
R.D. Gonzalez, Appl. Catal. A: Gen. 32 (1987) 185.
[40] M. Nawdali, D. Bianchi, Appl. Catal. A: Gen. 231 (2002) 45.
[41] S. Balcon, S. Mary, C. Kappenstein, E. Gengembre, Appl. Catal. A:
Gen. 196 (2000) 179.
[42] V. Mazzieri, F. Coloma-Pascual, A. Arcoya, P.C. Lárgentière, N.S. Fí-
goli, Appl. Surf. Sci. 210 (2003) 222.
[17] H. Ensslinn, M. Hartmann, L. Panizzon, US 2152960, Firm Soc.
Chemical Industry, Basel, Switzerland (1939).
[18] R.R. Fraser, R.B. Swingle, Can. J. Chem. 48 (1970) 2065.
[19] E. Ferber, H. Brückner, Chem. Ber. 72 (1939) 995.
[20] I. Fernández-Morales, A. Guerrero-Ruíz, F.J. López-Garzón, I. Rodrí-
guez-Ramos, C. Moreno-Castilla, Appl. Catal. 14 (1985) 159.
[43] J. Sárkány, M. Bartók, J. Catal. 92 (1985) 388.
[44] D.I. Kondarides, Z. Zhang, X. Verikios, J. Catal. 176 (1998) 536.
[45] S. Scire, S. Minic, C. Crisafulli, Appl. Catal. A: Gen. 235 (2002)
21.
[46] U.K. Singh, M.A. Vannice, Appl. Catal. A: Gen. 213 (2001) 1.
[47] L. Cerveny, Catalytic Hydrogenation, Elsevier, Amsterdam, 1986.