Multiphase catalytic hydrogenation of p-chloroacetophenone and acetophenone. A kinetic study of the reaction selectivity toward the reduction of different functional groups

Article abstract of DOI:10.1006/jcat.2000.3054

A multiphase catalytic system consisting of hydrocarbon solvent and aqueous phase, phase-transfer (PT) agent (Aliquat 336), and supported Pt or Pd catalyst, bubbled with hydrogen at atmospheric pressure, affords the rapid and efficient hydrodechlorination and the selective reduction of the functional groups of p-chloroacetophenone and acetophenone at 50°C and in moderate reaction times. The reaction gives quantitative yields of reduction products with variable selectivity, which can be controlled by varying some simple reaction conditions. Herein both halogen removal and carbonyl group or phenyl ring reductions are discussed. The composition of the aqueous phase, the nature of supported catalysts, the effect of various inorganic anions, and other reaction conditions have been studied kinetically to estimate the mechanistic aspects of the reaction selectivity. Emphasis is placed on the development of a novel synthetic tool, which allows control over the selectivity in the reduction of the carbonyl group and the phenyl ring. A number of kinetic models and the corresponding integral rate expressions were used, coupled with the simultaneous nonlinear least-squares fitting method, to estimate the rate constants for the formation of each particular reaction component. The kinetic monitoring allowed the accurate analysis of the reaction selectivity. Some mechanistic conclusions were drawn for the reaction under study that suggest that softer reduction conditions are attained in the presence of a PT agent and the aqueous phase.

Full text of DOI:10.1006/jcat.2000.3054

Journal of Catalysis 196, 330–338 (2000)
doi:10.1006/jcat.2000.3054, available online at http://www.idealibrary.com on
Multiphase Catalytic Hydrogenation of p-Chloroacetophenone and
Acetophenone. A Kinetic Study of the Reaction Selectivity
toward the Reduction of Different Functional Groups
Pietro Tundo,1 Sergei Zinovyev,2 and Alvise Perosa
Dipartimento di Scienze Ambientali, Universita` Ca’ Foscari, Dorsoduro 2137, 30123 Venezia, Italy
Received April 25, 2000; revised August 4, 2000; accepted August 31, 2000
rapid and efficient removal of chlorine and other halogens
(1–3) as well as the reduction of other functional groups
(3–5). The reaction gives quantitative yields of reduction
products under extremely mild conditions: hydrogen at at-
mospheric pressure, 50 C, and moderate reaction times,
whereas in most other systems known for HDH in the pres-
ence of supported metal catalysts (e.g., the gaseous phase
reaction under continuous flow), harsh temperature and
pressure are required (6–9). Moreover, the supported cata-
lyst under these conditions is easily poisoned by side prod-
ucts(e.g., HCl) (7, 10) and hasshort-lived activity. Examples
of HDH with polymer-immobilized Rh complexes are also
known (11), but involve more complicated systems.
A peculiarity of the reaction is its broad scope: in partic-
ular, it has synthetic value, and it is a promising technique
for the degradation of toxic chlorinated pollutants. From
the synthetic standpoint, it is a versatile reduction tech-
nique for the preparation of valuable products, and it is
conceivable that this may be done in an ecofriendly manner,
starting from the selective removal of halogen from unde-
sired chlorinated aromatics (12). There are, in fact, a vari-
ety of publications (13–16) on the catalytic heterogeneous
hydrodechlorination (HDC) over supported (Pt, Pd, Ru,
etc.) catalysts, aimed at the utilization of hazardous com-
pounds such as polychlorinated biphenyls (PCBs) (2), pes-
ticides (14), chemical weapons (15, 16), and other polychlo-
rinated aliphatics/aromatics, which have been produced in
huge amounts on the industrial scale.
A multiphase catalytic system consisting of hydrocarbon sol-
vent and aqueous phase, phase-transfer (PT) agent (Aliquat 336),
and supported Pt or Pd catalyst, bubbled with hydrogen at at-
mospheric pressure, affords the rapid and efficient hydrodechlo-
rination and the selective reduction of the functional groups of
p-chloroacetophenone and acetophenone at 50 C and in moder-
ate reaction times. The reaction gives quantitative yields of reduc-
tion products with variable selectivity, which can be controlled by
varying some simple reaction conditions. Herein both halogen re-
moval and carbonyl group or phenyl ring reductions are discussed.
The composition of the aqueous phase, the nature of supported
catalysts, the effect of various inorganic anions, and other reac-
tion conditions have been studied kinetically to estimate the mech-
anistic aspects of the reaction selectivity. Emphasis is placed on
the development of a novel synthetic tool, which allows control
over the selectivity in the reduction of the carbonyl group and the
phenyl ring. A number of kinetic models and the corresponding
integral rate expressions were used, coupled with the “simultane-
ous nonlinear least-squares fitting” method, to estimate the rate
constants for the formation of each particular reaction component.
The kinetic monitoring allowed the accurate analysis of the reac-
tion selectivity. Some mechanistic conclusions were drawn for the
reaction under study that suggest that “softer” reduction condi-
tions are attained in the presence of a PT agent and the aqueous
c
phase.
2000 Academic Press
Key Words: catalytic hydrogenation; multiphase conditions;
Pt/C catalyst; phase-transfer catalysis; kinetic study.
Multiphase HDH can be also used as a mild method for
the detoxification of hazardous chlorinated aromatic pollu-
tants and of persistent organic pollutants such as polychlori-
nated dibenzo-p-dioxins (PCDDs), polychlorinated diben-
zofurans (PCDFs), and PCBs (2). It can therefore become a
valuable technique for the detoxification of polychlorinated
pollutants present in organic wastes, incinerator gases, con-
taminated materials, and environmental matrices.
INTRODUCTION
It has recently been reported that a new hydrodehalo-
genation (HDH) system, consisting of aqueous basic and
organic phases, and a supported metal (Pt or Pd) cata-
lyst coupled with a phase-transfer (PT) agent, affords the
1 To whom correspondence should be addressed. Fax: +39 (041) 257
8620. E-mail: tundop@unive.it.
It has also been shown that under these new multiphase
conditions, which were originally developed for the dehalo-
genation of various haloaromatics, the selective reduction
2 On leave from N.D. Zelinsky Institute of Organic Chemistry of Rus-
sian Academy of Sciences, Leninsky pr. 47, Moscow 117913, Russia.
330
0021-9517/00 $35.00
c
Copyright
2000 by Academic Press
All rights of reproduction in any form reserved.

MULTIPHASE CATALYTIC HYDROGENATION
331
(4, 5) of carbonyl or hydroxyl group, or phenyl ring reduc- tector coupled to a HP 5890 gas chromatograph fitted with
tion, may take place as well. Yet, it has been shown (4) that a 30-m
by varying simple reaction conditions one can switch the
0.25-mm DB5 capillary column.
selectivity of the reduction toward more or less reduced
products or control the selectivity in the parallel reduction
steps, i.e., in the reduction of the carbonyl versus the aro-
matic ring. The main factors, which have been proven to
have such an influence, are the composition of the aque-
ous phase, and the nature and the ratios of the supported
metal catalyst and of the PT agent (4). Although the sys-
tem may appear complicated due to the number of different
phases present (we can count five: aqueous, organic, sup-
ported catalyst, PT agent which coats the catalyst and forms
a third liquid phase, and gaseous H2), and although inter-
pretation of how each variable comes into play is difficult,
the reaction itself is carried out easily even on preparative
scales.
The complete mechanism is still under investigation;
however, evidence suggests that the PT agent forms a third
liquid phase over the surface of the supported metal cata-
lyst, thereby modifying the chemical environment on the
solid catalyst surface. According to a model mechanistic
scheme published elsewhere (5), the reaction takes place in
this third liquid phase and the selectivity is affected by the
different geometry of substrate adsorption on the catalyst
surface. Therefore, the selective reduction of different func-
tional groups should be affected by the nature of the third
liquid phase, and such factors as the nature of PT agents,
the additions of various anions, and the pH of the aqueous
phase appear to be critical.
This work is aimed at the investigation of the reaction
selectivity, by measuring the observed kinetic constants
of each step. The kinetic data analysis and the detailed
monitoring of the different reaction conditions lead to
accurate conclusions about the peculiarities of the mul-
tiphase conditions studied. The examples chosen are p-
chloroacetophenone and acetophenone; the latter has been
studied in more detail since we were interested in the se-
lectivity of the reduction of functional groups other than
chlorine, and especially in the selectivity toward the reduc-
tion of the aromatic ring, which is unfavorable under the
usual catalytic conditions (6–9).
General Procedure for p-Chloroacetophenone and
Acetophenone Reduction
A 25-ml three-necked round-bottomed flask thermo-
stated at the reaction temperature (50 0.1 C), connected
with a system for the addition of hydrogen and with a water-
jacketed condenser, was employed for the reaction. If not
otherwise indicated, the reactor was loaded with a mix-
ture of an aqueous phase (5.7 ml; for the composition of
the aqueous phase see the conditions indicated in the ta-
bles), 5% Pt/C (0.0837 g; 0.021 mmol of Pt), Aliquat 336
(tricaprylmethylammonium chloride, 0.103 g, 0.26 mmol),
0.7 mmol of substrate (p-chloroacetophenone or acetophe-
none), and n-decane as the internal standard (0.0407 g,
0.3 mmol). The organic phase was adjusted to 10 ml by
adding the corresponding amount of isooctane. The reac-
tion mixture was magnetically stirred at 1000 rpm. Hydro-
gen was bubbled at atmospheric pressure into the organic
phase at about 5 ml/min.
Sampling Procedure and GC Analyses
The samples were collected from the organic phase dur-
ing the reaction at time intervals. The first one was taken
just after the reagents had been loaded and it was accounted
for as zero time. About 20 l of the reaction mixture was
taken for each sample, which was then diluted with ethyl
ether up to 1–2 ml. The resulting sample was then shaken
with silica and filtered with cotton wool to eliminate the
impurities of Aliquat 336 and Pt/C particles.
Determination of the concentration of the reaction com-
ponents was performed using the internal standard tech-
nique. For this reason, a number of calibration coefficients
were calculated for each reaction component by analyzing
the binary mixtures of these compounds with n-decane. The
calibration coefficients were calculated by plotting the ra-
tios of the component–decane GC peak areas versus the
ratios of their known concentrations in the corresponding
binary mixtures. Taking into account the known concentra-
tion of the internal standard, the calibration coefficients
(kcal) for p-chloroacetophenone, acetophenone, phenyl
ethanol, cyclohexyl ethanol, acetyl cyclohexane, ethyl ben-
zene, ethyl cyclohexane, p-chlorophenylethanol, and p-
METHODS
All reagents and solvents were ACS grade and were chloroethylbenzene were calculated as 0.046, 0.039, 0.038,
used without further purification. The 5% Pt/C was from 0.040, 0.037, 0.026, 0.035, 0.046, and 0.031, respectively. The
Fluka, Art. No. 80982; its surface area was 700–800 m2/g. following expression was used for the calculation of compo-
The 10% Pd/C was from Aldrich, Art. No. 20, 569-9. GC nent concentrations: Ccomp = kcal(Scomp/Sstand), where Scomp
analyses were performed on a Varian GC 3400 using a and Sstand are the GC peak area units for the analyzed
fused silica capillary column “Chrompack CP Sil 8 CB” (30 compound and for the internal standard, respectively. The
m
0.32 mm) with 95% dimethyl–5% diphenyl polysilox- results were successfully tested for reproducibility. The sys-
ane as the stationary liquid phase (film thickness 0.25 m). tematic error for reproduction, reflected in the determi-
GC/MS analyses were performed on a HP 5971 mass de- nation of rate constants, was no higher than 10% . The

332
TUNDO, ZINOVYEV, AND PEROSA
first-order rate constants were obtained by using a fitting where ki j is the intrinsic rate constant for the i j-th step, Pi
technique described herein.
RESULTS AND DISCUSSION
and PH are the partial pressures of the i-th reactant and
2
hydrogen, respectively, the term in the denominator is the
surface term, and Ki is the equilibrium adsorption coeffi-
cient. The surface term will cancel since the surface is iden-
tical for all species, and assuming that the partial pressure of
hydroge0n is constant, then the observed relative rate con-
stant ki j (if the usual first-order approach is used) will be
represented as a product of the intrinsic rate constant ki j
and the equilibrium adsorption constant Ki. This approach
has been used for the similar case of heterogeneous HDH
in the vapor phase (9).
However, even if the adsorption coefficients do affect
the observed rate constants, we still assume that the ad-
sorption corrections can be omitted: in fact, the present ki-
netic model is mainly directed toward the appraisal of the
actual reaction selectivity, and it is applicable to this end.
The intrinsic rate constants cannot be easily calculated since
the determination of the adsorption coefficients is trou-
blesome; under our multiphase conditions the adsorption
terms (Ki and the surface term) appear to be different for
each intermediate from one system to another, since the
catalyst surface is modified by the components, according
to the reported mechanism (5). Moreover, the multiphase
system seems too complicated for the usual adsorption es-
timation (17), because the reactants do not directly sorb on
the catalyst surface, but transfer to the different environ-
ment of the third liquid layer.
Observed Rate Constants and Adsorption Contribution
The investigation of the hydrogenation reaction un-
der multiphase conditions has been carried out using p-
chloroacetophenone and acetophenone as model com-
pounds. Since the reaction pathways usually differ from one
system to another, it is helpful to use a scheme which shows
all the possible reaction pathways (Scheme 1) for both p-
chloroacetophenone and acetophenone reductions. Steps
which account for the reduction of the aromatic ring leav-
ing chlorine did not occur in any system and therefore are
not included.
Each reaction step in this scheme has its own rate con-
stant. The observed kinetics are rather complicated due to
there being a number of intermediates, but can be described
using the developed mathematical approach if each step is
assumed to be pseudo-first-order (see the following para-
graph).
Sets of calculated first-order rate constants for each
particular experiment have been obtained using the “si-
multaneous nonlinear least-squares fitting” technique and
are listed in tables using the definitions corresponding to
Scheme 1 (kn(0), n = 1, 2, . . . , 7). The rate constants obtained
remain the observed ones since adsorption coefficients are
not considered. Differential adsorption of all the reaction
intermediates and reagents over the supported catalyst is
an important factor which contributes to the observed rate
constants. According to the Langmuir–Hinshelwood ap-
proach, exemplified by Beranek (17) for some complex ki-
netic schemes in heterogeneous catalysis, the reaction rate
can be expressed as
Besides affecting rate constants, adsorption may affect
the reaction kinetics as well, by causing deviation of the
observed reaction profiles from the model first-order ki-
netics. The error due to this effect and to side products can
be corrected by verifying the mass balance of the reaction.
In our case, the reaction components can be accounted for
within 10% , implying that the reaction kinetics can still be
accurately processed using the first-order approach.
ki j Ki P Pn(m)
i
Kinetic Models and Data Analysis
H2
ri j
=
P
b
1 +
Ki P
,
i
A number of kinetic models and the corresponding math-
ematical equations have been developed in order to per-
form the computational integral kinetic analysis of the data
obtained. The mathematical approach for Scheme 1 ap-
pears to be too complicated to perform an accurate esti-
mation of the experimental data. Moreover, in all the cases
studied, depending on the conditions employed, only some
particular reaction steps take place, and so only the inter-
mediates corresponding to these steps have been analyzed.
Therefore, the kinetic approach employed for the particu-
lar cases makes it possible to reduce the overall reaction
scheme to a number of less complicated ones. All the re-
action pathways, which take place in the reductions of ace-
tophenone and p-chloroacetophenone, can be divided into
seven different cases; for each of them the corresponding
mathematical approach has been developed. The global
mechanistic scheme is presented in Scheme 7 of Table 1.
i
SCHEME 1. Possible reaction pathways for the catalytic hydrogenation
of p-chloroacetophenone and acetophenone.

MULTIPHASE CATALYTIC HYDROGENATION
333
TABLE 1
the following set of integral rate expressions:
Model Mechanistic Schemes Used for the Interpretation of
Reaction Pathways
(k1+k2)t
[A] = [A]0e
[1]
[2]
k1
(k1+k2)t
(k3+k5)t
e
[B] = [A]0
e
k5 + k3 K2 k1
k2
(k1+k2)t
k4t
[C] = [A]0
e
e
[3]
k4 k1 k2
k1k3
1
[D] =
k2 +
[A]0
k1 + k2
k3 + k5
1
(k1+k2)t
(k3+k5)t
1
e
e
k5 + k3 k2 K1
k2
(k1+k2)t
k4t
(k1+k2)t
e
e
e
[4]
[5]
k4 k2 k1
k1 k5
1
[E] =
[A]0 1
k1 + k2 k3 + k5
k5 + k3 k2 k1
(k1+k2)t
(k3+k5)t
(k3 + k5)e
(k1 + k2)e
.
These expressions, which are derived for the most complex
mechanistic scheme (entry 7 in Table 1), can then be easily
reduced to onescorrespondingto allthe other applied cases.
To processthe experimentaldata, the “simultaneousnon-
linear least-squares fitting” technique has been performed
by means of the “MicroCal Origin v. 5.0 PC software using
the “simplex” iterations method. This technique allows a
number of integral rate expressions (e.g., Eqs. [1]–[5]), de-
scribing the formation (decomposition) of all the reaction
components (e.g., A–E) in each particular experiment, to
be used together to process all the corresponding data sets
(the sets of experimental data on the component concen-
trations) simultaneously.
The simultaneous curve fitting appears to be an accurate
technique for the analysis of complex reaction schemes.
For example, it allows one to consider such complex ex-
pressions as [4] and [5], which contain four or five param-
eters and therefore are not likely to be applied separately.
The simultaneous fitting technique affords a set of rate con-
stants for each experiment by minimizing the approxima-
tion errors with respect to the formation of each compo-
nent, which depends on these constants. An example of the
“nonlinear least-squares fitting” procedure is presented in
Fig. 1.
The definitions used for compounds (A, B, C, D, E) and
for rate constants (k1, k2, k3, k4, k5) in Table 1 represent the
formal approach and do not necessarily correspond to those
in Scheme 1. These definitions can be substituted for ones
which are used for the actual constants by simply matching
the geometry of the reaction pathway with the correspond-
ing one in Table 1.
Based on the model mechanistic schemes presented in
Table 1, a number of integral kinetic expressions can be
written. Each of these expressions accounts for a certain
reaction intermediate, product, or substrate. It is obvious
that the most complex model scheme (entry 7, Table 1) can
be reduced to all the other cases (entries 1–6) by eliminating
certain reaction steps. The same is true for the mathemat-
ical expressions if the corresponding rate constants are set
to zero. The integral rate expressions, which represent the
concentration dependencies of A, B, C, D, and E for entry
7, are derived from the following system of differential rate
expressions:
d[A]
dt
d[B]
dt
=
(k1 + k2)[A],
= k1[A] (k3 + k5)[B],
= k3[B] + k4[C],
d[C]
dt
d[D]
dt
Sets of first-order rate constants obtained for each ex-
periment and the corresponding standard approximation
errors are presented in Tables 2–5 using the models of
Table 1. The choice of the model was based on the pre-
liminary appraisal of the component kinetic profiles (18).
To understand which model has been used, it is enough to
examine which constants are presented in the tables. The
standard errors reported rarely exceed 20% , which is quite
acceptable for the integral kinetic analysis.
= k2[A] k4[C],
d[E]
= k5[B].
dt
By setting the initial conditions [B], [C], [D], [E] =
0, [A] = [A]0 at t = 0, and the mass balance [A]0 =
[A] + [B] + [C] + [D] + [E], the solution for this system is

334
TUNDO, ZINOVYEV, AND PEROSA
biphase aqueous–organic system, even in the absence of
Aliquat 336 and base (Table 2, entry 1). The reaction pro-
ceeds exclusively through the 1 → 2 → 3 → 6 → 9 pathway,
affording ethyl cyclohexane (9) as a major product (90% )
with a comparatively high reaction rate. Small amounts of
dechlorinated ketone (4) and alcohol (5) have been de-
0
tected, which are formed with relatively low rates (k2
k1 , k3
0
0
0
k4 ) and readily give acetyl cyclohexane (7) and
cyclohexyl ethanol, the latter in only 5–7% yield.
The results suggest that under these conditions chlorine
is not likely to be removed from the aromatic ring before
the other functional groups are reduced. The prohibition of
the 8 → 9 step is also interesting and may result from the
effect of the acidic medium, since the HCl released during
the dechlorination may cause leaching processes (7) and
inhibition of the activation of H2 on the catalyst (10), or
from competing adsorption of aromatic intermediates on
the catalyst. Seemingly, the aliphatic (cyclohexyl) deriva-
tives are not likely to undergo reduction in the presence
of aqueous acid. This observation is also supported by the
hydrogenation of acetophenone in the presence of aqueous
acid and Aliquat 336 (Table 3).
FIG. 1. “Nonlinear least-squares fitting” procedure on the example of
hydrogenation of acetophenone on Pt/C with Aliquat 336 and pure water
(Table 3, entry 2). The kinetic model employed is that of entry 5 in Table 1.
Dots and curves represent the experimental data and the corresponding
theoretical integral rate dependencies, respectively.
When any of the intermediates or products are formed
in negligibly small amounts (e.g., ethyl benzene and ethyl
cyclohexane in Scheme 1), their exclusion from the ki-
netic scheme is warranted, as it simplifies the mathemat-
ical model, and it reduces the error on the other constants.
Moreover, since the rate constants of their formation can-
not be accurately measured, only the initial rate of forma-
tion isestimated bycalculatingthe tg at time = 0. The same
linear approximation for the initial rate of formation was
used for the products of very slow reaction steps.
Hydrogenation of p-Chloroacetophenone with
Aliquat 336 and Base
In the presence of Aliquat 336 and of 1% aqueous KOH
solution, the hydrogenation of p-chloroacetophenone pro-
ceeds with reversed selectivity (Table 2, entry 2). Chlorine
is the first functional group to be reduced, so no formation
of 2 and 3 is observed. The reaction pathway follows the
1 → 4 → (7 + 5) order, with the final selectivity toward the
formation of 5. Only negligible rates for the formation of 6,
8, and 9 have been observed. Under these multiphase con-
ditions, there is a high consecutive selectivity toward ace-
tophenone (4) because the rate for the removal of chlorine
(1 → 4) is much higher than those of all subsequent reaction
Hydrogenation of p-Chloroacetophenone without Base
and without PT Agent
p-Chloroacetophenone can be easily dechlorinated and steps (4 → (7 + 5)). This same effect has been previously
reduced to ethyl cyclohexane with H2 and Pt/C in a observed for p-chloropropiophenone which, in the absence
TABLE 2
First-Order Rate Constants for the Hydrogenation of p-Chloroacetophenone over 5% Pt/C
1
First-order constants 103 (s
)
Products
molar ratio
No.
k1
k2
k3, k4, k5
k6
k7
k10
k20
k30
k40
k50
c
c
c
d
d
e
1a
2b
—
—
—
0
0
0
0.58 0.13
0
1.45 0.18
0
—
—
0
—
0
2.10 0.38
0
1/15 (8/9) f
1/10 (7/5)g
0.17 0.02
0.02 0.00
5.68 0.85
a No Aliquat 336 in the presence of water.
b Aliquat 336 and aqueous KOH (1.0 mmol).
c Considering negligible amounts of 4, 5, and 7, k1 5 cannot be determined accurately.
d (k20 + k3 ) 0.05(k10 + k4 ).
0
0
e k40 cannot be determined accurately, and is estimated as k10 < k40 < k5 .
0
f After 80 min.
g After 3 h.

MULTIPHASE CATALYTIC HYDROGENATION
335
TABLE 3
First-Order Rate Constants for the Hydrogenation of Acetophenone over 5% Pt/C under Multiphase Conditions in the Presence
of Neutral, Basic, and Various Acidic Aqueous Phases
1
First-order constants 103 (s
)
No.
Composition of aqueous phase
k1
k2
k3
k4
k5
k6
k7
1
2
3
4
5
6
Basic medium, KOH (1.0 mmol) 0.25 0.01
0
0
0
0
0
0
0
0
0
—
—
Neutral, water
0.17 0.01 0.13 0.01 0.05 0.01 0.03 0.01
6 a
5 a
7 a
7 a
b
Slightly acidic, 0.003 M HCl
Acidic, 0.118 M HCl
Strongly acidic, 1.18 M HCl
Acidic,c 0.118 M HCl
0.20 0.01 0.12 0.01 0.10 0.01
0.40 0.02 0.12 0.01 0.16 0.02
1.38 0.04 0.15 0.01 0.13 0.01
0
0
0
(7 0.7) 10
(2 0.2) 10
0.48 0.02
(1 0.05) 10
(3 0.1) 10
b
0
0
0.07 0.01
0.60 0.15
1.45 0.17 0.45 0.13 0.37 0.07 0.30 0.17
0.27 0.03
a Initial rates (M/s).
b Not determined.
c No Aliquat 336.
of the PT agent but with 1% KOH, or vice versa with the ring (4 → (5 + 7) → 8 pathway), so the rate constants for
PT agent (A336) and pure water, also initially underwent these steps (k1 and k2) are comparable. As 5 and 7 are
dechlorination (4, 5).
formed, they undergo further, but slower, reduction to 8,
These results can be compared to those obtained for the which is the final reduction product. Among all the other
hydrogenation of acetophenone in the presence of Aliquat conditions tested, the use of water, i.e., the neutral medium,
336 and 1% aqueous KOH (Table 3, entry 1). In the latter results in the most favorable selectivity (nearly 40% ) to-
case no ring reduction product (7) is observed and the rate ward the reduction of the phenyl ring while in most other
constant for the formation of 5 (k1) is somewhat higher. This cases it is shifted toward 5.
difference probably arises from the decrease in the concen-
The presence of a basic aqueous phase (1% aqueous
tration of KOH (initially 1.0 mmol) during the reaction of KOH, entry 1) results in rather mild reduction. Only the
p-chloroacetophenone, where it is partially neutralized by 4 → 5 step takes place; i.e., the selectivity is completely
the HCl (0.7 mmol) being released.
shifted to the reduction of the carbonyl group and 5 is the
only reaction product. The rate constant for this step (k1) is
slightly higher than that in neutral medium or in the system
where p-chloroacetophenone is used as a substrate (Table 2,
entry 2).
Mutiphase Hydrogenation of Acetophenone in the
Presence of Acidic/Basic Aqueous Phase
It has recently been reported (5) for the hydrogena-
tion of p-chloropropiophenone that the composition of the
aqueous phase has a substantial effect on the selectivity in
the catalytic multiphase reduction of different functional
groups. One of the crucial factors appears to be the acid-
ity/basicity of the aqueous phase.
With an aim to perform a detailed kinetic study of the se-
lectivity of the reduction of other (than chlorine) functional
groups, namely the carbonyl, and phenyl ring reduction and
to clarify the conditions that affect the reaction kinetics and
govern the selectivity, a series of experiments have been
carried out. To simplify the model reaction schemes and
to eliminate side effects which may arise from the removal
of chlorine, all the forthcoming results have been obtained
starting from the catalytic hydrogenation of acetophenone.
A number of systems, which employ neutral, basic, or var-
ious acidic compositions of aqueous phase, were probed.
The rate constants for the reaction steps which occur in
these systems and the corresponding conditions are re-
ported in Table 3.
Switching to acidic conditions (entries 3–5) the selectiv-
ity of the first step is again on the side of carbonyl reduction
of the aryl ketone (while the aliphatic ketone is unreactive
as explained earlier). The reaction pathway preferably fol-
lows the 4 → (5 → 7), 5 → (6 + 8), 8 → 9 order in moderate
acidic media and the 4 → (5 + 7), 5 → (6 + 8), 6 → 9 order
if more acidic conditions are employed. Figure 2 displays
the change in rate constants k1, k2, k3, k4 versus the concen-
tration of aqueous HCl. While the rate constants for the
other steps are apparently constant and comparable, there
is a sharp increase in the selectivity toward the reduction of
carbonyl group with an increase in the HCl concentration.
Increasing the acidity of the aqueous phase, there is also an
increase in the yields of final reduction products, namely of
ethyl benzene (6) and ethyl cyclohexane (9), especially if
the strong acidic medium is employed (entry 5). This trend
is in agreement with the fact that carbonyl hydrogenoly-
sis is favored over hydrogenation in the presence of HCl
(19).
Remarkably, the presence of aqueous acid results in the
complete prohibition of the 7 → 8 and 8 → 9 steps (k4, k6 = 0
in entries 3–5). The observed reaction pathway follows the
The use of pure water at pH = 7 (entry 2) affords the
equally fast reduction of both carbonyl group and phenyl

336
TUNDO, ZINOVYEV, AND PEROSA
if Aliquat 336 and an acidic environment are present can
be again explained by considering the weaker adsorption
of the aliphatic substrates, especially if the surface is deac-
tivated by the presence of HCl. The observed phenomenon
seems to be a combined effect of HCl and Aliquat 336,
which cause a weaker adsorption of the aliphatic species.
Indeed, if only acidic (in the absence of Aliquat 336, Table 3,
entry 6) conditions or if only Aliquat 336 (neutral medium,
Table 3, entry 2) is introduced in the reaction mixture the
step corresponding to the reduction of aliphatic ketone (7)
is equally prohibited. The 8 → 9 step is prohibited in the
presence either Aliquat 336 or acid, or under most other
monitored conditions, but it is more likely to be prohibited
due to the deactivation of the Pt catalyst over time. Presum-
ably, the absence of PT agent produces harsher conditions
for the reduction, so that the 7 → 8 step takes place.
FIG. 2. Effect of acidity on the kinetics of acetophenone hydrogena-
tion over 5% Pt/C under multiphase conditions.
Influence of Aqueous Anions on the Reaction Kinetics
The additions of various inorganic salts have also been
shown to affect the reaction rates. It was expected that
lipophilic anions would substitute chlorine in Aliquat 336
and therefore change the selectivity, since the nature of
the PT agent has a major influence on the selectivity. The
rate constants for the corresponding reaction steps in the
systems where different salts were added are reported in
Table 4.
It was found that all the tested salts can be divided into
two major groups. While the additions of most salts result in
more or less pronounced effects on the reaction kinetics, KI,
Na2S, and KSCN kill the catalyst. Only the initial rate was
measured to estimate the reaction kinetics in these cases.
The majority of the other salts tested, which are listed in
Table 4, also shift the selectivity of the reaction toward the
4 → (5 + 7) → 8, 5 → 6 → 9 sequence. Nevertheless, if
acetophenone is reduced in acidic medium and in the ab-
sence of Aliquat 336 (entry 6) the 7 → 8 step takes place,
though it is quite slow and there is still no sign of the 8 →
9 step. The same is true (k6 = 0) if p-chloroacetophenone
is reduced in the absence of Aliquat 336 with a neutral
aqueous phase, where HCl is formed (Table 2, entry 1).
Moreover, the rate constants for the 6 → 9 step (k7) in
the reduction of both p-chloroacetophenone (Table 2, en-
try 1) and acetophenone (Table 3, entry 6) are perfectly
reproduced, showing that the mathematical approach is sat-
isfactory. The “acidic medium” behavior in the case of p-
chloroacetophenone is due to the release of HCl during the
HDC. The fact that the 7→ 8 and 8 → 9 steps are prohibited
TABLE 4
First-Order Rate Constants for the Hydrogenation of Acetophenone over 5% Pt/C under Multiphase Conditions (Aliquat
336, Aqueous Phase) in the Presence of Salt Additions
1
First-order constants 103 (s
)
Salt
(0.26 mmol)
k1
k2
k3
k4
k5
k6
KCla
0.23 0.01
0.19 0.01
0.42 0.02
0.16 0.01
0.15 0.01
0.14 0.01
0.07 0.00
0.10 0.01
0.02 0.00
0
0
0
0
0
KBra
KBrb
KI
Na2S
KSCN
CH3COONa
KHSO4
NaSO4C12H23
NaSO4C12H23
0.08 0.00
0
4
10
2
10
2
10
3
10
5
6 d
5
7 d
7
8 d
8 d
8 d
7
8 d
9 d
9 d
7
8 d
7
2
3
10
10
10
5
3
8
10
10
10
3
7
7
10
10
10
2
6
5
10
10
10
1
1
1
10
10
10
1
5
1
10
10
10
7
7
8
8
7
7
9 d
9 d
8
7
7
7
9 d
8 d
9 d
8 d
8
7
7
7
9 d
8 d
9 d
9 d
0.12 0.01
0.37 0.02
0.33 0.02
0.84 0.07
0.07 0.00
0.17 0.01
0.12 0.02
0.20 0.04
0.02 0.00 0.06 0.01
5
3
1
3
10
10
10
10
6
3
8
2
10
10
10
10
8
3
1
3
10
10
10
10
3
1
5
6
10
10
10
10
0.09 0.01
0.09 0.01
0.17 0.01
0
0.10 0.01
0.04 0.02
c
a 0.7 mmol of salt.
b 0.7 mmol of salt in the presence of aqueous HCl (0.118 M).
c In the absence of Aliquat 336.
d Initial rates (M/s).

MULTIPHASE CATALYTIC HYDROGENATION
337
One of the most interesting effects observed is the sub-
stantial acceleration of the reaction rates in the presence of
a long-chain lipophilic anion (dodecylsulfate), especially if
it is used in the absence of Aliquat 336. Probably, such salts,
which manifest the properties of surfactants, may produce
a co-catalyzing micellar effect by modifying the surface of
a supported Pt catalyst. This effect does not seem to be
the same as the one produced by the PT agent, because
Aliquat 336 does not usually speed up the reaction, but
rather softens the conditions affording the yields of less re-
duced products. The effect of surfactant in this case affords
a faster reduction at all steps, not much affecting the se-
lectivity. Seemingly, this effect is suppressed if PT agent is
already sorbed on the surface.
FIG. 3. Effect of KI aqueous concentration on the initial rates in the
multiphase hydrogenation of acetophenone over 5% Pt/C.
Hydrogenation of Acetophenone under Other Conditions
To extend the scope of the monitored conditions, which
may affect the behavior of the catalytic hydrogenation, and
to elucidate the peculiarities of the multiphase systems, a
series of experiments set up in alternative systems has also
been carried out. The results on the rate constants for these
systems are reported in Table 5.
As was previously mentioned, the absence of Aliquat 336
results in much harsher reduction conditions. Under these
conditions, much higher rate constants for all the reaction
steps are observed, especially for the formation of the final
reduction products (6, 9). The selectivity of the first step is
shifted to the formation of 5.
In the absence of aqueous phase, there is no substantial
difference in the rate constants compared to the those of
blank reaction (Table 3, entry 2). The selectivity on the
first step is slightly shifted to the formation of 5, and like
in acidic media, the reduction of acetyl cyclohexane (7) is
suppressed. There are also small amounts of final reduction
products (6 and 9) formed.
reduction of the carbonyl group. Compared to the blank ex-
periment, where no salt is added (pure water as the aque-
ous phase, entry 2 in Table 3), this effect is more or less
pronounced from one salt to another and results mainly
from the acceleration of k1, while k2 and k3 apparently stay
constant. The exception is CH3COONa, which results in a
two-fold decrease in the rate constants for all the reaction
steps comparing to the blank reaction, but k1 still prevails
over k2. Moreover, in the presence of some salts the overall
reduction is deeper, so that the higher yields of the final
reduction products (6 and 9) are observed.
Additions of salts such as KHSO4 or KBr have an effect
similar to that in the acidic media, namely the complete
inhibition of the 7 → 8 step (k4 = 0). This is understandable
for KHSO4, which is acidic itself, but is quite surprising for
KBr. The use of KBr in the presence of aqueous acid leads
to a further increase in k1 (it is the same as without KBr,
but only acid is added), while all the other rate constants
are not affected.
A two-fold decrease in the amount of Aliquat 336 brings
no essential changes to the reaction kinetics, though it
TABLE 5
First-Order Rate Constants for the Hydrogenation of Acetophenone in Various Catalytic Systems
1
Conditions
First-order constants 103 (s
)
No. Catalyst
Aq. phase PT agent
Solvent
k1
k2
k3
k4
k5
k6
1
2
3
4
5
6
5% Pt/C
5% Pt/C
5% Pt/C
5% b Pt/C
5% Pt/C
10% Pd/C
Water
None
Water
Water
None
Water
None
Aliquat
Isooctane 0.64 0.01 0.24 0.09 0.42 0.10 0.34 0.22
Isooctane 0.26 0.01 0.13 0.01 0.08 0.01
0.14 0.03
0.53 0.10d
7
8 c
9 c
7
9 c
9 c
0
2
1
10
3
8
10
1
1
10
10
6
1
10
10
Aliquata Isooctane 0.17 0.01 0.14 0.01 0.08 0.01 0.02 0.00
10
10
10
10
7
7
Aliquat
None
Aliquat
Isooctane 0.06 0.00 0.04 0.00 0.02 0.00 0.02 0.01
0
0
0
0
7
8 c
7
8 c
Ethanol
0.52 0.02 0.08 0.01 0.02 0.00
0
0
8
1
2
10
1
10
Isooctane 0.68 0.05 0.02 0.00
0
a Halved amount of Aliquat 336 is used (0.13 mmol).
b Halved amount of Pt/C is used (0.0419 g).
c Initial rates (M/s).
d k7.

338
TUNDO, ZINOVYEV, AND PEROSA
causes some reduction toward the final hydrogenation proceedstoward the formation ofcompletelyreduced prod-
products. Nevertheless, even this halved amount of Aliquat ucts. Moreover, the acidic multiphase conditions prohibit
336 produces substantial changes compared to the reac- the reduction of cyclohexyl derivatives.
tion with no PT agent. Probably, there is saturation on the
amount of Aliquat 336, which is required to produce the may have a substantial impact on the reaction kinetics. Salts
effect of multiphase conditions. such as iodides, sulfides, or thiocyanides inhibit the reduc-
It has also been found that additions of inorganic salts
Remarkably, a two-fold decrease in the amount of sup- tion by poisoning the surface of the supported metal cata-
ported Pt catalyst leads to the corresponding decrease in lyst, while the majority of other inorganic salts do not ex-
all the reaction rates, not much affecting the selectivity. This ert any pronounced influence. Remarkably, in the presence
testifies that only the supported metal catalyst contributes of surfactants such as sodium laurylsulfate the reaction is
to the actual catalytic effect, and the hydrogenation reac- significantly accelerated, especially if no PT agent is used.
tion is not possible in the absence of it.
This reveals new possibilities to modify the multiphase
Under otherwise similar conditions, 10% Pd/C sharply conditions by coupling them with a co-catalyzing micellar
differs from 5% Pt/C, resulting in a much milder (not deep) effect.
reduction. The use of the Pd catalyst resembles the behavior
ACKNOWLEDGMENTS
of the Pt catalyst in the presence of aqueous base, where
only one-step reduction takes place with selectivity strongly
shifted to the reduction of the carbonyl group.
The reaction in ethanol in the absence of Aliquat 336 and
aqueous phase constitutes the usual heterogeneous cata-
lytic hydrogenation. The rate constants for most steps are
rather low except for a high value for k1. Remarkably, the
“acidic-like” effect takes place again in this case (k4 = 0).
This work was supported by INCA (Interuniversity Consortium Chem-
istry for the Environment) and MURST (Ministero Universita´ e Ricerca
Scientifica e Technologica) fondo 60% .
REFERENCES
1. Marques, C. A., Selva, M., and Tundo, P., J. Chem. Soc., Perkin Trans.
1 529 (1993); Marques, C. A., Selva, M., and Tundo, P., J. Org. Chem.
58, 5256 (1993); Marques, C. A., Rogozhnikova, O., Selva, M., and
Tundo, P., J. Mol. Catal. A: Chemical 96, 301 (1995).
2. Marques, C. A., Selva, M., and Tundo, P., J. Org. Chem. 59, 3830
(1994).
CONCLUSIONS
In this work, a kinetic study has been carried out to es-
timate the rates and the selectivities and to determine the
effects of different reaction conditions on the selectivity in
this multiphase catalytic hydrogenation system. A kinetic
approach and a number of mathematical rate expressions
are derived and successfully applied for the estimation of
rate constants at each reaction step. The possible effect
of the reagent/intermediate adsorption over the supported
catalyst does not much affect the observed first-order ki-
netics and so the adsorption corrections were not included
in the observed rate constants.
3. Marques, C. A., Selva, M., and Tundo, P., J. Org. Chem. 60, 2430
(1995).
4. Selva, M., Tundo, P., and Perosa, A., J. Org. Chem. 63, 3266 (1998).
5. Perosa, A., Selva, M., and Tundo, P., J. Org. Chem. 64, 3934 (1999).
6. Okano, T., Tsukiyama, K., and Konishi, H., Chem. Lett., 603–606
(1982).
7. Hoke, J. H., Gramiccioni, G. A., and Balko, E. N., Appl. Catal: B 1,
285–296 (1992).
8. Shin, E.-J., and Kaane, M. A., Catal. Lett. 58, 141–145 (1999).
9. La Pierre, R. B., Wu, D., Kranich, W. L., and Weiss, A. H., J. Catal.
52, 59–71 (1978).
10. Pasett, P., and Gentile, F., J. Org. Chem. 60, 1351 (1995).
11. Okano, T., Tsukiyama, K., Konishi, H., and Kiji, J., Chem. Lett., 603
(1982); Tung, H.-S., Brubaker, C. H., Jr., J. Organomet. Chem. 216,
129 (1981).
It has been shown that the use of multiphase conditions,
namely the use of a PT agent coupled with a supported
metal catalyst in the presence of aqueous phase, results in 12. Bomben, A., Marques, C. A., Selva, M., and Tundo, P., Synthesis
1109 (1996).
the selective acceleration of the chlorine removal as well
as in softer reduction conditions, preventing the formation
of the final reduction products and giving substantial yields
of either aromatic ring or carbonyl reduction products. The
13. Musoke, J., Roberts, D., and Cooke, M., Bull. Environ. Contam.
Toxicol. 28, 467 (1982).
14. “Pesticides Disposal and Detoxification: Processes and Techniques”
(A. P. Dillon, Ed.), Noyes Data Corp., Parkridge, NJ, 1981.
selectivity in the first step can be controlled to some extent 15. Hagh, B. F., and Allen, D. T., in “Innovative Hazardous Waste
Treatment Technology Series” (H. M. Freeman, Ed.), Vol. 2, Physical/
Chemical Processes, pp. 45–53. Technomic Publ. Inc., Lancaster, PA,
1990.
and is mostly shifted to the reduction of the carbonyl group.
The most favorable selectivity toward reduction of phenyl
ring is observed if no KOH is used as an aqueous phase
16. Kalnes, T. N., and Jemmes, R. B., in “Innovative Hazardous Waste
in the presence of Aliquat 336. The completely selective
reduction of carbonyl occurs in the presence of aqueous
base or if Pd catalysts are used.
The effects of basic and acidic aqueous media have been
investigated. The use of base results in the selective reac-
tion toward the carbonyl group reduction, whereas in acidic
media the reduction conditions are harsher so the reaction
Treatment Technology Series” (H. M. Freeman, Ed.), Vol. 1, Thermal
Processes, pp. 65–76. Technomic Publ. Inc., Lancaster, PA, 1990.
17. Beranek, L., in “Advances in Catalysis,” Vol. 24, pp. 1–55. Academic
Press, San Diego, CA, 1975.
18. Schmid, R., and Sapunov, V. N., “Non-Formal Kinetics.” Verlag
Chemie GmbH, Weinheim, 1982.
19. Armandia, M. A., Borau, V., Jimenez, C., Marinas, J. M., Sempere,
M. E., and Urbano, P., Appl. Catal. 43, 47 (1988).