Activity enhancement by the support in the hydrogenation of C=C bonds over polymer-supported palladium catalysts

Article abstract of DOI:10.1002/1521-3765(20000602)6:11<1980::AID-CHEM1980>3.0.CO;2-Z

Four synthetic ion-exchange resins (AH, BH, CH, DH) of different hydrophilic/hydrophobic properties were used as supports for heterogeneous palladium catalysts (A, B, C, D). The resins contained styrene (STY) and 2-(methacryloxy)ethylsulfonic acid (MESA)

Full text of DOI:10.1002/1521-3765(20000602)6:11<1980::AID-CHEM1980>3.0.CO;2-Z

FULL PAPER
Activity Enhancement by the Support in the Hydrogenation of
ˆ
C C Bonds over Polymer-Supported Palladium Catalysts
Marco Zecca,*[a] Roman FisÏera,[b] Giancarlo Palma,[c] Silvano Lora,[d]
Milan Hronec,[b] and Milan KraÂlik[b]
Abstract: Four synthetic ion-exchange
resins (AH, BH, CH, DH) of different
the third comonomer. The catalysts (Pd
0.25 ± 0.45% w/w) were obtained by ion-
exchanging the acidic forms of the resins
with [Pd(OAc)2] and reducing palladi-
um(ii) with excess sodiumborohydride.
The use of NaBH4 also ensured the
neutralization of the acidic sites of the
supports. No effect of the hydrophilic/
hydrophobic properties of the supports
was observed in the hydrogenation of
cyclohexene and 2-cyclohexen-1-one in
methanol, at 258C and 0.5, 1, and
1.5 MPa, respectively. However, cata-
lysts A and B, containing amido groups
provided by either DMAA or MBAA,
proved to be more active than C and D.
The observed activity enhancement was
directly proportional to the nitrogen/
palladiummolar ratio in the catalysts.
This finding suggests that amido groups
promote palladium through a direct
interaction with the metal surface.
hydrophilic/hydrophobic
properties
were used as supports for heterogeneous
palladiumcatalysts (A, B, C, D). The
resins contained styrene (STY) and
2-(methacryloxy)ethylsulfonic
acid
(MESA) as the comonomers. Either
divinylbenzene (DVB: CH, DH resins)
or
N,N'-methylenebisacrylamide
(MBAA: AH, BH resins) were used as
the cross-linker. AH contained also
N,N-dimethylacrylamide (DMAA) as
Keywords: hydrogenations ´ ion-ex-
change resins ´ palladiumcatalysts
synthesis of alkylmethyl ethers and of bisphenol A.[4, 5] More-
over, heterogeneous palladiumcatalysts supported onto ion-
exchange resins are currently applied to the chemical removal
Introduction
Polymer-based catalysts have been under investigation for at
least three decades. Academic research in this field has been
mainly focused on the polymer-bound counterpart of homo-
genous transition metal catalysts (hybrid catalysts),[1±3] but no
industrial accomplishment has been achieved yet. By contrast,
functionalized cross-linked polymers are currently exploited
as true heterogeneous catalysts in a number of industrial
processes. In particular, ion-exchange resins based on sulfo-
nated polystyrene are employed as acidic catalysts in the
[6]
of dioxygen fromwater, down to the ppb level, and to the
one-pot synthesis of methylisobutylketone from acetone.[5]
It is well known that the polymeric supports can affect the
progress of the catalytic process in several ways. First, the
local concentration of chemical species inside polymer-based
catalysts can be different fromthe phase of reactants and
products, as the consequence of kinetic, mass transport, and
thermodynamic (partition between the catalyst and the
reactants/products phases) effects. Thus, the affinity of the
reagents and/or products towards the polymeric support plays
an important role, not only in polymer-bound phase transfer
catalysis,[7] but also in other cases. For instance, in the partial
hydrogenation of benzene over rutheniumcatalysts supported
onto both inorganic and polymeric carriers,[8, 9] the selectivity
depends on the hydrophilic/hydrophobic properties of the
supports. Of course, the affinity of reagents and/or products
towards the catalytic support depends on the chemical
structure of the latter and it must be emphasized that
polymeric materials can be much more easily tailored than
inorganic solids under this respect. Moreover, in the case of
polymer-supported metal catalysts, the functional groups of
polymeric support could also interact directly with the active
[a] Dr. M. Zecca
Dipartimento di Chimica Inorganica, Metallorganica e Analitica
Universita di Padova, via Marzolo 1, 35131 Padova (Italy)
Â
Fax : (‡39)-049-8275223
E-mail: mzecca@chin.unipd.it
Ï
Â
[b] Dr. R. Fisera, Prof. M. Hronec, Prof. M. Kralik
Department of Organic Technology
Slovak University of Technology
Radlinskeho 9, 812 37 Bratislava (Slovakia)
Â
[c] Prof. G. Palma
Â
Dipartimento di Chimica Fisica, Universita di Padova, via Loredan 2,
35131 Padova (Italy)
[d] Dr. S. Lora
Istituto di Fotochimica e Radiazioni di Alta Energia
C.N.R., via Romea 4, 35020 Legnaro (Padova) (Italy)
1980
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0611-1980 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 11

1980±1986
component and affect the intrinsic kinetics of the reaction.
However, whereas the influence of the polymeric support on
the catalytic activity has been studied in some detail for
polymer-anchored metal complexes,[10] relatively little atten-
tion has been paid to this issue in connection with polymer-
supported metal catalysis.[11]
Results and Discussion
Commercially available sulfonated polystyrene/divinyl-ben-
zene ion-exchange resins were employed in our previous work
on the hydrogenation of 4-nitrotoluene.[16] These materials are
usually fully sulfonated and therefore are highly hydrophilic.
At the start of this work we aimed at the preparation of
polymer-based catalysts with controlled hydrophilic/hydro-
phobic properties, as we were interested in ascertaining
whether cyclohexene and 2-cyclohexen-1-one showed differ-
ent affinities towards the supports and how the reaction was
eventually affected. Therefore, we prepared four copolymers
(AH, BH, CH, DH) of 2-(methacryloxy)ethylsulfonic acid
(MESA), which served as the source of both sulfonic groups
for ion-exchange and hydrophilic monomeric units, with
styrene (STY) which supplied the hydrophobic monomeric
units. AH contained also a substantial proportion of a third,
hydrophilic comonomer, N,N-dimethylacrylamide (DMAA).
The resins were obtained in their acidic forms. Previous
experience showed that i) cross-linked copolymers of MESA
and STY synthesized in N,N-dimethylformamide (DMF)
solution contain substantial amounts of both the solvent and
water,[14] and ii) whereas STY copolymerizes almost quanti-
tatively, MESA does not. This also occurred in the synthesis of
AH, BH, CH, and DH, as confirmed by their elemental
analyses (C, H, N, S, Table 1). The elemental analyses,
evaluated in a similar way as in ref. [13], revealed that the
newly formed resins contained 3 ± 10% (w/w) of DMF and 5 ±
10% (w/w) of water. The relevant data are reported in Table 1
and Table 2 and the chemical structure of the resins is
sketched in Scheme 1.
For a few years we have been investigating polymer-
supported metal catalysts.[12±17] Our attention was mainly
focused on the activity and stability of palladiumnano-
particles dispersed inside gel-type resins. We studied the
hydrogenation of cyclohexene,[15] of benzene,[9] and, in greater
detail, the hydrogenation of aromatic nitrocompounds.[16, 17] In
this context, particular attention was paid to the deactivation
of the catalyst. Investigation of the molecular mobility inside
swollen polymeric catalytic supports and of their morphology
at the nanometric scale provided a sound basis for the
interpretation of the behavior of polymer-based catalysts.[14, 15]
Then, our interest shifted to the interactions between the
substrate, the support, and the catalytic sites. Due to our
previous experience concerning the stability of polymer-based
catalysts in the hydrogenation of cyclohexene, we decided to
study this reaction with catalysts of different hydrophilic/
hydrophobic properties. Moreover, we chose a second sub-
strate of different hydrophilic/hydrophobic features, 2-cyclo-
hexen-1-one, and also studied its hydrogenation on the same
catalysts. The hydrogenation of a double carbon ± carbon
bond conjugated to a keto group is not only of academic
interest, but has also a practical relevance: For instance, the
hydrogenation of mesityl oxide (4-methyl-3-pentene-2-one) is
the second step of the industrial one-pot synthesis of
methylisobutylketone, which is carried out over polymer-
supported bifunctional palladiumcatalysts.
The solvent compatibility of the four resins, which depends
on their hydrophilic/hydrophobic properties, was assessed
fromthe bulk expanded volumes values in several solvents,
which are reported in Figure 1. Comparable values of bulk
expanded volumes were also obtained for the catalysts after
We report herein on our results on the hydrogenation of
cyclohexene and 2-cyclohexen-1-one, which was found to be
insensitive to the hydrophilic/hydrophobic properties of the
supports. However, reaction rates appeared to be enhanced
by the presence of amido groups in the polymeric carriers.
Table 1. Compositions (weight%)[a] of resins AH, BH, CH, and DH.
AH
BH
CH
DH
calcd found calcd found calcd found calcd found
1
2
3
4
5
6
% C
% H
% N
% S
% water
% DMF
54.37 54.87 69.29 68.68 57.41 56.97 72.02 72.83
Abstract in Italian: Quattro resine per scambio ionico (AH,
7.78 7.52
5.46 5.43
4.95 4.97
10
7.22 7.07
1.61 1.51
4.52 4.45
4.8
6.98 6.64
0.53 0.51
7.36 7.38
9.8
7.37 7.37
0.58 0.58
4.02 3.99
6.0
Á
BH, CH, DH), dotate di proprieta idrofiliche/fobiche diffe-
renti, furono utilizzate quali supporti per catalizzatori etero-
genei di palladio (A, B, C, D). I catalizzatori furono ottenuti
mediante immobilizzazione di palladio(ii) per scambio ionico
a partire da [Pd(OAc)2] e sua successiva riduzione con NaBH4
in eccesso. Fu quindi studiata la reazione di idrogenazione del
10
4.8
3.2
3.6
[a] The expected elemental weight percentages were calculated on the
assumption that the resins contained water and DMF in the proportion
reported in entries 5 and 6 and that polymerization yields were as reported
in Table 2.
À
doppio legame C C nel cicloesene, nel cicloesanone e nel
2-cicloesen-1-one, a 258C e alla pressione di 0.5, 1.0, e 1.5 MPa,
rispettivamente, in presenza dei catalizzatori A, B, C, D. Alcuni
dei supporti polimerici contenevano gruppi ammidici e i
catalizzatori basati su di essi risultarono pi‚ attivi degli altri,
indipendentemente dalla pressione di idrogeno. Inoltre fu
Table 2. Polymerization yields (mol%) for each comonomer in the
synthesis of resins AH ± DH.
AH
BH
CH
DH
Á
osservato che lꢀaumento di attivita catalitica era direttamente
STY
MESA
DVB
DMAA
MBAA
90
85
±
100
100
100
85
±
±
100
100
80
100
±
100
80
100
±
Á
proporzionale al rapporto molare fra azoto e palladio. Cio
suggerisce un effetto di promozione diretta del catalizzatore da
parte dei gruppi ammidici eventualmente presenti nel supporto.
±
±
Chem. Eur. J. 2000, 6, No. 11
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0611-1981 $ 17.50+.50/0
1981

M. Zecca et al.
FULL PAPER
As shown in Figure 2, the BEV value of all investigated
resins was much higher in pure methanol than in both pure
cyclohexene and cyclohexanone, which gave generally negli-
gible swelling of the polymer network, if any. This is probably
À
due to the fact that whereas in methanol the SO3H groups
are extensively dissociated, in cyclohexene or cyclohexanone
ionic pairs are strongly held together. Thus, in methanol the
electrostatic repulsion among the fixed negative ionic charges
helps swelling. However, the BEV of BH in methanol was the
highest for this resin, but it was practically negligible in water
(Figure 1). This indicates that in general the electrostatic
repulsion among fixed charges is not the only factor aiding
swelling. Thus, methanol must also be able to interact
considerably with the hydrophobic domains, different from
water. This highlights the amphiphilic nature of methanol.
Resin DH showed the same behavior upon swelling in
methanol/cyclohexene and methanol/cyclohexanone mixtures
of different compositions. In fact, the changes of BEV as a
function of the molar fraction of methanol (cOH) were
practically the same in both cases. The BEV of DH in pure
methanol was the lowest among the investigated resins and,
moreover, it was very poorly swollen by both pure cyclo-
hexene and cyclohexanone. This highlights the hydrophobic
nature of DH and the generally poor swelling power of
cyclohexene and cyclohexanone. When cOH was increased
from0 up to about 0.25, the BEV reached practically the same
value as in pure methanol. Further increase of cOH did not
further change the BEV value. This suggests that for DH the
interaction of methanol, cyclohexene, and cyclohexanone are
relatively weak hence the resin always takes up selectively the
better swelling agent available, that is methanol.
Scheme 1. Chemical structures of resins AH, BH, CH, DH.
Figure 1. Bulk expanded volumes (BEVs) of resins AH, BH, CH, and DH.
reduction with NaBH4. The data in Figure 1 indicate that
hydrophilicity increases in the following order:
BH  DH < CH
ꢀ
AH
(clearly
hydrophilic)
(moderately
(moderately
hydrophilic)
hydrophobic)
The behavior of CH swollen with methanol/cyclohexene
mixtures was qualitatively the same. However, as BEV in pure
methanol is higher than for DH, the initial increase was
steeper for CH and furthermore the BEV increased almost
linearly. These findings point to the same direction as before:
Due to very weak interaction of cyclohexene with the resin,
This sequence is in agreement with the expectation that the
lower the proportion of STY in the resin, the higher its
hydrophilicity.
After the synthesis, the resins were exchanged with
palladium(ii), which was subsequently reduced to the metal
with excess NaBH4 in ethanol.
This reduction procedure en-
sures a homogeneous metal
distribution throughout the
volume of the catalyst beads
and the removal of DMF ad-
sorbed in the support.[13]
In order to get insight about
the relative affinities of cyclo-
hexene and 2-cyclohexen-1-
one, we measured the bulk
expanded volumes (BEVs) of
resins AH, BH, CH, and DH
swollen with mixtures of meth-
anol with cyclohexene and
cyclohexanone, respectively,
at different compositions. Cy-
clohexanone was taken as a
mimic of the unsaturated cy-
clic ketone and used instead.
The results are illustrated in
Figure 2.
Figure 2. Bulk expanded volumes of catalysts A, B, C, and D in cyclohexene/methanol mixtures (squares) and
2-cyclohexanone/methanol mixtures (diamonds) versus mixturesꢁ composition.
1982
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0611-1982 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 11

Hydrogenation
1980±1986
Table 3. Mole fractions (%, c) of methanol, before[a] and after[b] swelling of
catalysts A, B, C, and D with a 1m methanol solution in a) cyclohexene,
b) cyclohexanone, and c) 2-cyclohexen-1-one.
methanol was absorbed preferentially to allow the largest
possible swelling of the polymer network.
By contrast, the behavior of CH swollen with methanol/
cyclohexanone mixtures was totally different. When cOH was
progressively increased, a practically linear steady growth of
the BEV value was observed in the whole composition range
(cOH ˆ 0 ± 1). This indicates that some specific interaction
occurs between CH and cyclohexanone. Moreover, the BEV
in pure cyclohexanone (cOH ˆ 0) was considerably higher than
in pure cyclohexene, which also suggests that the cyclic ketone
has actually a higher affinity towards CH than the cyclic alkene.
When resins AH and BH were swollen with methanol/
cyclohexene mixtures, their behavior resembled that of CH
and DH. However, the BEV did not reach the value observed
in pure methanol at cOH as low as 0.25 ± 0.30. After that point a
further linear, albeit smoother, increase of the BEV values
was recorded up to cOH ˆ 1. This suggests that the cyclohexene
interaction with AH and BH could be slightly stronger than
with CH and DH.
mass of catalyst
(dry, mg)
mass of methanol
(mg)
c after
before
after
a) cyclohexene
A
B
C
D
107.6
100.0
98.3
34.6
32.1
37.8
35.0
9.7
9.4
10.9
10.1
3.3
3.3
2.8
2.1
102.4
b) cyclohexanone
A
B
C
D
118.9
109.1
103.0
96.2
31.2
36.0
31.2
40.8
9.1
10.7
9.5
6.3
7.2
6.9
0.7
12.1
c) 2-cyclohexen-1-one
A
B
C
D
101.4
105.6
103.7
101.2
32.3
37.0
33.0
41.1
12.2
12.3
10.2
13.1
3.7
3.2
3.1
0.9
[a] Theoretically calculated from the employed amounts of methanol and
solvent; [b] measured in the liquid phase by GC (see also Experimental
Section).
On the other hand, when resins AH and BH were swollen
with methanol/cyclohexanone mixtures, their behavior was
qualitatively the same as that of CH. Also in this case the data
indicate that cyclohexanone has a higher affinity than cyclo-
hexene towards AH and BH, respectively.
for catalyst D the decrease in cOH for the cyclohexanone
solution was not very different fromthose observed for the
cyclohexene and 2-cyclohexen-1-one solutions. This indicates
that the relative affinities of cyclohexene and cyclohexanone
(and 2-cyclohexen-1-one) towards the support of catalyst D
are similar, in agreement with the BEV data discussed above.
Moreover, for each kind of solution the decrease in cOH for
catalyst D was always larger than for the other materials.
These facts are also consistent with the hypothesis that the
partition process is mainly driven by swelling in the absence of
strong interactions of methanol, cyclohexene, cyclohexanone,
and 2-cyclohexen-1-one with the support.
For all the other catalysts (A, B, C) the decrease in cOH for
the cyclohexanone solution was always smaller than in the
cyclohexene solution. This finding indicates that cyclohex-
anone competes ªfor the polymerº with methanol better than
cyclohexene, which is also fully coherent with the hypothesis
that cyclohexanone has a higher affinity than cyclohexene
towards the polymeric supports of catalysts A ± C. The
decrease in cOH values observed for the methanol solution
in 2-cyclohexen-1-one were comparable with those found for
solutions of methanol in cyclohexene, thus indicating a
relatively poor affinity of 2-cyclohexen-1-one towards the
investigated materials. Incidentally, this finding shows that
cyclohexanone is not an appropriate mimic of 2-cyclohexen-1-
one.
Further information concerning methanol ± support sub-
strate affinity was obtained fromthe comparison of solute
concentrations, before and after swelling, in methanol/cyclo-
hexene and methanol/cyclohexanone mixtures used to swell
the catalysts. As described in the experimental, they were
actually swollen in a known amount of ªsolventº and, after
equilibration, a known amount of ªsoluteº was added to the
suspension (in the following discussion ªsolventº and ªsoluteº
are the most and the least abundant component, respectively,
of the swelling mixture, see also Experimental Section). The
molar fraction of the ªsoluteº before swelling was not
measured directly, but it was calculated as described in the
Experimental Section. On the other hand, molar fractions
after swelling were directly measured in the liquid phase in
equilibriumwith the swollen polymer.
We carried out two sets of experiments, where methanol
played the role of either the ªsoluteº (Table 3) or the
ªsolventº (Table 4).
When methanol was used as the ªsoluteº, it was added to
the resins previously swollen in cyclohexene, cyclohexanone,
or 2-cyclohexen-1-one. Its addition generally brought about
an appreciable increase of the BEV of the support, as it was
expected from the BEV measurements. The molar fraction of
methanol (cOH) in the free liquid phase was always found
significantly lower after swelling than before, that is the
partition coefficient of methanol is always much greater than
one, although it was not possible to calculate its value from
our data.[18] This clearly indicates that, in general, methanol
was preferentially driven into the polymer network, probably
as the consequence of stronger methanol ± support interaction
with respect to the ªsolventº-support interaction. This finding
is completely coherent with BEV values found for the resins
in methanol, cyclohexene, and cyclohexanone.
The data of Table 4 were obtained when methanol was used
as the ªsolventº. In this case, the molar fractions of the
ªsoluteº (cyclohexene, cyclohexanone, and 2-cyclohexen-1-
one) in the free liquid phase in equilibriumwith the swollen
resins were practically the same as before swelling. This is
again in agreement with the hypothesis that methanol
interacts much more strongly with the supports than the
cyclic alkene and ketones. In the presence of a large excess of
the alcohol, BEVs were very close or equal to those observed
in pure methanol. This was apparently the case when the
As for the relative affinity of cyclohexene and cyclohex-
anone for the various materials, the data of Table 3 show that
Chem. Eur. J. 2000, 6, No. 11
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0611-1983 $ 17.50+.50/0
1983

M. Zecca et al.
FULL PAPER
concentration of cyclohexene was 1m (c ˆ ca. 4  10À2; Fig-
ure 2), the same as in the catalytic tests discussed below. This
indicates that under these conditions the polymer chains are
fully solvated by methanol molecules. This prevents the
interaction of the ªsoluteº with the polymeric support,
independently of its nature. Hence, it can almost freely flow
fromthe liquid phase to the swollen polymer and its partition
coefficient approaches unity.
Our data on the partition of cyclohexene and 2-cyclohexen-
1-one demonstrate that both has similar affinities towards
each of the employed catalysts and therefore we need not
worry about this possibility. Moreover, the results described
above show that at a 1m concentration in methanol, even
substrates of different affinity are distributed to the same
extent between the liquid and swollen polymeric phase, due to
the levelling effect of very strong methanol ± support inter-
actions. As for metal leaching, we did not observe loss of the
metal in any case investigated.
An appreciable decrease, albeit not large, of the molar
fraction of both cyclohexanone and 2-cyclohexen-1-one was
observed for catalyst C (Table 4), which is based on the
The initial relative hydrogenation rates at 0.5, 1.0, and
1.5 MPa for both substrates are reported in Table 5 (the values
Table 4. Mole fractions (%, c) of a) cyclohexene, b) cyclohexanone, and
c) 2-cyclohexen-1-one before[a] and after[b] swelling of resin-based catalysts
A, B, C, D with a 1m solution in methanol.
Table 5. Initial reaction rates (molsÀ1 kgPdÀ1) and relative initial reaction
rates (in brackets) with respect to the rate over the catalysts C, for
cyclohexene (ene) and 2-cyclohexen-1-one (one).
mass of catalyst mass of a), b), or c), respectively
c
(dry, mg)
(mg)
before after
Catalyst
Substrate
1.5 MPa
a) cyclohexene
ene
one
A
B
C
D
98.4
104.2
104.0
98.0
130
155
142
157
5.6
6.4
5.1
4.5
5.6
6.3
5.1
4.6
0.5 MPa 1 MPa
0.5 MPa 1 MPa
1.5 MPa
A
B
C
D
80 (2.1) 102 (2.2) 114 (2.3)
52 (1.4) 68 (1.5) 76 (1.5)
84 (2.1) 104 (2.1) 117 (2.6)
54 (1.4) 64 (1.3) 72 (1.4)
b) cyclohexanone
38 (1)
46 (1)
55 (1)
40 (1)
49 (1)
54 (1)
A
B
C
D
99.4
104.3
102.0
83.6
137
135
152
145
5.4
4.9
5.8
5.3
5.2
5.0
5.1
5.4
44 (1.2) 52 (1.1) 58 (1.2)
46 (1.2) 53 (1.1) 57 (1.2)
were calculated fromthe first derivative of the quadratic
polynomial fitting of points from 0 up to 70% of the
conversion). The data show that the hydrogenation rates of
both substrates change to a very similar extent from a catalyst
to another.
c) 2-cyclohexen-1-one
A
B
C
D
102.2
101.6
101.8
97.0
149
157
152
145
5.6
5.7
5.3
5.2
5.5
5.6
4.9
5.1
In addition, if the ratios of the initial hydrogenation rates of
2-cyclohexen-1-one and cyclohexene (Table 6) are compared
for each catalyst at different pressures, it appears that the
substrate which is hydrogenated faster depends on the
pressure of hydrogen. All these data confirmthat the hydro-
philic/hydrophobic properties of the support do not play an
important role.
[a] Theoretically calculated from the employed amounts of methanol and
solvent; [b] measured in the liquid phase by GC (see also Experimental
Section).
support with the highest content of sulfonic groups. This
finding suggests that the main interaction of cyclohexanone
and 2-cyclohexen-1-one with supports generally involves the
ionic sites of the latter, with effects appearing only at
relatively high content of the sulfonic groups. The particular
affinity of cyclohexanone towards resin CH, highlighted by
the relatively high BEV value of this material in the pure
ketone, also supports this idea.
Table 6. Initial rate over catalysts A, B, C, and D of 2-cyclohexen-1-one
hydrogenation relative to cyclohexene.
Cat
0.5 MPa
1 MPa
1.5 MPa
A
B
C
D
1.05
1.04
1.05
1.05
1.02
0.94
1.07
1.02
1.03
0.95
0.98
0.98
The catalytic tests were carried out under kinetic regime, as
checked with catalysts particles of different sizes (0.1 ±
0.14 mm and 0.14 ± 0.18 mm). Hence, mass transport phenom-
ena are not rate determining in the tests described herein.
Therefore, the effects of swelling, on which diffusion in
polymeric materials dramatically depends, can be neglected.
Under kinetic control, the concentration of the involved
species (reactants, products) is homogenous throughout the
catalyst and depends only on the respective affinities, which
depend in turn on the hydrophilic/hydrophobic properties of
the supports. Accordingly, they could be expected to affect the
concentration of reactants and products thermodynamically
rather than kinetically. In fact, if two substrates are adsorbed
to different extents, rather than at different rates, onto a
supported catalyst, different rates of the catalytic process
could be eventually observed because of their different
concentration at the active sites, even assuming that the
catalyst is equally active for both.
Catalyst B appears to be generally a little more active than
D. The most important difference between them is the nature
of the cross-linker: Whereas D contains divinylbenzene, B
contains methylene-bis-acrylamide. Moreover, catalyst A,
which incorporates a substantial amount of N,N-dimethyla-
crylamide, is by far the most active one, hence we can argue
that increasing the amount of nitrogen in the resins enhances
the catalyst activity.
In fact, the rate enhancement can be quantitatively
correlated to the molar ratio between nitrogen and palladium
in the catalysts. The true nitrogen content, that is without the
contribution of DMF, which is removed during the reduction
of palladium(ii), can be obtained fromthe elemental analyses
1984
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0611-1984 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 11

Hydrogenation
1980±1986
data of Table 1. If the initial hydrogenation rates of cyclo-
hexene and 2-cyclohexen-1-one over catalysts A, B, and C
(the latter assumed as a typical unpromoted catalyst) are
plotted as a function of the respective nitrogen/palladium
molar ratios, the curves depicted in Figure 3 are obtained.
Figure 4. Schematic representation of the interaction of the surface of a
palladiumcrystallite with the amido groups fromthe polymeric support.
2-cyclohexen-1-one at 0.5, 1.0, and 1.5 MPa in methanol at
258C. On the one hand, it was observed that when the
catalysts were added to the reaction mixtures (1m solutions of
the substrates), only methanol interacted with the polymeric
chain of the resins and this quenched any substrate ± support
interaction. On the other hand, we get some evidence that
anyway the substrates have similar affinities towards the
catalyst resins.
We have serendipitously found that the presence of amido
groups in the polymeric supports enhances the rate of the
chemical reaction. Under the employed conditions the rate
enhancement is proportional to the molar ratio between
nitrogen in the support and the active metal. This result
indicates that it is possible to ªplayº with the functional
groups of polymeric carriers of polymer-supported metal
catalysts not only to match their physico-chemical properties
to the nature of solvent, reagents, and products, but also to
change directly the intrinsic reactivity of the active metal.
Figure 3. Initial hydrogenation rate of cyclohexene a) and 2-cyclohexen-1-
one b) over catalysts A, B and C versus nitrogen/palladium(N/Pd) molar
ratio in the catalysts.
Although some care is necessary in drawing conclusions
(each curve contains only three points), the results are
apparently the same at all of the investigated pressure values:
Hydrogenation rates are linearly dependent on the N/Pd ratio
in the catalysts for both substrates. The direct proportionality
of the reaction rates to N/Pd ratios indicates that nitrogen
fromthe amido groups of DMAA and MBAA directly
promotes the palladium catalyst. Moreover, amido groups
fromDMAA and fromMBAA are comparably effective in
promoting the catalysts, in that the only source of amido
groups in catalyst B is the cross-linker.
Experimental Section
Since mechanistic details of the reaction are not known yet,
we cannot decide whether the effects of this interaction are
mainly electronic or steric. However, according to the theory
of catalysis over metals and alloys,[19] electronic effects seem
to be dominant. We have already argued that the promoting
effect depends on a direct interaction of amido groups with
the palladiumcrystallites. We can further speculate that a
donor± acceptor interaction between the nitrogen atoms,
which possess free electron pairs, and palladiumatoms of
the surface likely occurs. A similar ªcoordinativeº interaction
between the surface atoms of a polymer-protected platinum
colloid and the nitrogen atoms of amido groups from the
protecting poly(N-isopropylacrylamide) matrix has been
recently proposed.[11b] This could build up electron density
on the metal, thus making a process like the dissociative
adsorption of dihydrogen easier (Figure 4).
Materials and methods: Cyclohexene, methanol (MeOH), ethanol
(EtOH), toluene (tol), acetone (Me2CO), tetrahydrofuran (THF), diethyl
ether (Et2O), and ethyl acetate (AcOEt) were supplied fromLachema,
Brno, Czech Republic. MBAA, DMAA, and DMF were obtained from
Aldrich. Reagent-grade MESA, STY, and DVB were supplied by Strem,
Newburyport (MA), USA. 2-Cyclohexen-1-one, sodiumborohydride, and
palladium(ii)acetate were purchased fromFluka. All the products were
employed as received, with the exception of cyclohexene and 2-cyclohexen-
1-one, which were purified prior to use by distillation. Atomic absorption
measurements for Pd determination were performed on a Carl Zeiss Jena
AAS 3 atomic absorption spectrometer.
Synthesis of polymers AH, BH, CH, DH: The polymers were prepared by
g-ray irradiation of DMF solutions of the monomers at room temperature,
as described in ref. [13]. The compositions of the polymerization mixtures
are given in Table 7.
Solvent compatibility of AH, BH, CH, DH: The solvent compatibility of
the resins was evaluated fromtheir bulk expanded volumes (BEVs) in
different solvents as described in ref. [13].
Table 7. Composition of the monomer mixtures for the synthesis of resins
AH, BH, CH, and DH.
Resins
Monomer[a]
DMAA
ST
MESA
MBAA
DVB
Conclusion
AH
BH
CH
DH
38.0
74.4
53.3
74.7
27.8
21.7
41.9
20.2
30.3
3.9
3.9
±
±
±
4.8
5.1
±
±
±
The differences in hydrophilic/hydrophobic properties of the
resins in the investigated resin-supported palladiumcatalysts
did not affect the hydrogenation rate of cyclohexene and
±
[a] mol%.
Chem. Eur. J. 2000, 6, No. 11
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0611-1985 $ 17.50+.50/0
1985

M. Zecca et al.
FULL PAPER
Preparation of catalyst A, B, C, D: The palladiumcatalysts supported on
polymers were prepared starting from the acid form of the parent resins by
exchanging them with the appropriate amount of palladium(ii) acetate in
1:1 acetone/methanol. Each product was washed with methanol and
reduced prior to use with 0.066m sodiumborohydride in ethanol (10 equiv)
for 1 h, then washed with methanol, and activated with hydrogen in
methanol at 258C and 0.5 MPa for 10 min in the hydrogenation reactor.
After the activation, the catalyst was washed three times with methanol and
immediately used for the catalytic tests. The analysis for palladium was
carried out as described previously,[17] and yielded the following metal
content (w/w, %): A: 0.460, B: 0.266, C: 0.362, D: 0.371
[1] G. Parshall, Homogeneous Catalysis, 1st ed., Wiley, New York, 1980,
p. 227.
[2] F. R. Hartley, Supported Metal Complexes, Reidel, Dordrecht, 1985.
[3] H. Hirai, N. Toshima in Tailored Metal Catalysts (Ed.: Y. Iwasawa),
Reidel, Dordrecht, 1986, p. 87.
[4] P. M. Lange, F. Martinola, S. Oeckl, Hydrocarbon Process 1985, 51.
[5] H. Widdecke in Synthesis and separations using functional polymers
(Eds.: D. C. Sherrington, P. Hodge), Wiley, Chichester, 1988, pp. 149 ±
179.
[6] R. Wagner, P. M. Lange, Erdöl, Erdgas, Kohle 1989, 105, 414.
[7] M. Tomoi, W. T. Ford in Synthesis and separations using functional
polymers (Eds.: D. C. Sherrington, P. Hodge), Wiley, Chichester, 1988,
p. 181.
Partition measurements: Catalysts A, B, C, D were swollen with mixtures
of methanol with cyclohexene, cyclohexanone, and 2-cyclohexen-1-one,
respectively. After equilibration, the molar fraction of the most dilute
species were measured in the liquid phase. Two sets of experiments were
carried out for each catalyst, which had been swollen with:
[8] a) J. Struijk, M. dꢁAngremond, W. J. M. Lucas-de Regt, J. J. F. Schol-
ten, Appl. Catal. A.: General 1992, 83, 263; b) J. Struijk, R. Moene, T.
van der Kamp, J. J. F. Scholten, Appl. Catal. A.: General 1992, 89, 77.
Ï
Â
[9] M. Hronec, Z. Cvengrosova, M. Kralik, G. Palma, B. Corain, J. Mol.
Catal. 1996, 105, 25.
i) solutions of methanol (ªsoluteº) in cyclohexene, cyclohexanone, and
2-cyclohexen-1-one (ªsolventº), respectively;
[10] P. Hodge in Synthesis and separations using functional polymers (Eds.:
D. C. Sherrington, P. Hodge), Wiley, Chichester, 1988, p. 43.
[11] a) P. E. Garrou, B. C. Gates in Synthesis and separations using
functional polymers (Eds.: D. C. Sherrington, P. Hodge), Wiley,
Chichester, 1988, p. 123; b) C.-W. Chen, M-Q. Chen, T. Serozawa,
M. Akashi, Chem. Commun. 1998, 831.
ii) solutions of cyclohexene, cyclohexanone, and 2-cyclohexen-1-one (ªsol-
uteº), respectively, in methanol (ªsolventº).
The dry catalyst (ca. 100 mg) was put into a weighed stoppered flask. After
weighing the flask with the solid within, a defined volume of ªsolventº (see
above) was poured into the flask, which was stoppered and weighed again.
The amount of added ªsolventº was slightly higher than the minimum
required for complete swelling. Thus, after deposition of the swollen
material only a little proportion of ªsolventº was present as free liquid. The
catalysts were let to swell for 12 h. Then, the ªsoluteº was added and the
flask was weighed again. The amount of ªsoluteº was such to set its
concentration, with respect to the overall volume of the ªsolventº, to about
1m. The exact amounts of the catalyst, ªsolventº, and ªsoluteº were
measured as the difference between consecutively measured weights. After
6 h of occasional stirring, the swollen catalyst was let to settle. Then,
samples of the liquid were taken for GC analysis after which the mole
fraction of the ªsoluteº was determined (molar fraction after swelling). In
view of the experimental procedure, it was not possible to measure the
molar fraction of the ªsoluteº before swelling. This was calculated from the
masses of ªsolventº and ªsoluteº.
Â
[12] M. Kralik, M. Hronec, S. Lora, G. Palma, M. Zecca, A. Biffis, B.
Corain, J. Mol. Catal.: A Chem. 1995, 97, 145 ± 155.
Â
[13] M. Kralik, M. Hronec, V. Jorík, S. Lora, G. Palma, M. Zecca, A. Biffis,
B. Corain, J. Mol. Catal.: A Chem. 1995, 101, 143 ± 152.
Â
[14] M. Zecca, M. Kralik, M. Boaro, G. Palma, S. Lora, M. Zancato, B.
Corain, J. Mol. Catal.: A Chem. 1998, 129, 27 ± 34.
Â
[15] M. Kralik, M. Zecca, P. Bianchin, A. DꢁArchivio, L. Galantini, B.
Corain, J. Mol. Catal.: A Chem. 1998, 130, 85 ± 93.
Ï
Â
Â
[16] R. Fisera, M. Kralik, J. Annus, V. Kratky, M. Zecca, M. Hronec,
Collect. Czech. Chem. Commun. 1997, 62, 1763 ± 1775.
Â
[17] M. Kralik, R. Fi„era, M. Zecca, A. A. DꢁArchivio, L. Galantini, K.
Ï Â
Jerabek, B. Corain, Collect. Czech. Chem. Commun. 1998, 63, 1074 ±
1088.
[18] Some intrinsic problems arise when the partition of a species occurs
between a liquid and a polymer which swells in contact with the liquid
itself. First, not only the overall volume of the polymer phase changes
upon swelling, but it also depends on the compositions of the liquid
phase, sometimes dramatically, as it is shown also in this paper. This
makes difficult to assess the volume of the polymer phase, in which
concentrations must be determined. In fact, BEV is not suitable,
because it also includes the interparticle volume, which does not
belong to the swollen polymer phase. Moreover, it is matter of debate
whether the concentration of solutes in the swollen polymer phase
must be calculated with respect to its overall volume or to the fraction
which is not occupied by the polymer chains, that is in the volume
available to solvent molecules. Also in this case, the assessment of this
volume in swellable polymers is not an easy task. It has not a fixed
value, like in rigid porous solids, but it depends on the swelling
properties of the polymer network. The volume fraction occupied by
the polymer chains is a (decreasing) function the swelling volume: The
larger the latter, the smaller the former.
Catalytic tests: The catalytic tests were performed in a 25 cm3 glass-lined
stainless steel reactor connected with a flexible metal capillary to an
apparatus for measuring the hydrogen consumption at constant pressure
similar to that described in ref. [13]. Typically, 6 cm3 of a 1m solution
cyclohexene or 2-cyclohexen-1-one in methanol were employed, with the
amount of catalyst required to get an analytical concentration of palladium
equal to 0.125 (1.5 MPa) or 0.25 (0.5 and 1 MPa) molmÀ3. The reactor was
loaded with the freshly activated catalyst, followed by the reactants, and
hydrogen, put in a thermostated oil bath and vigorously shaken at about
12 Hz. The end of hydrogen take-up was considered as the end point of the
reaction. The reaction products in the final reaction mixture were also
analyzed by GC to confirmthe total conversion of the substrate.
Acknowledgements
[19] V. Ponec, G. C. Bond, Catalysis by Metals and Alloys, Elsevier,
Amsterdam, 1995.
This work was supported by funds of the project SK-95/195/205:
Ecologically clean technologies for refinery, chemical and petrochemical
industry.
Received: October 7, 1999 [F2069]
1986
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0611-1986 $ 17.50+.50/0
Chem. Eur. J. 2000, 6, No. 11