Catalysis with palladium colloids supported in poly(acrylic acid)-grafted polyethylene and polystyrene

Article abstract of DOI:10.1139/V06-076

Grafts of poly(acrylic acid) on polyethylene powder (PE-g-PAA) or polystyrene (PS-g-PAA) can be used to support Pd(O) crystallites that function like a homogeneous Pd(O) catalyst in some reactions. These Pd-PE-g-PAA catalysts were active in allylic substitution reactions in the presence of added phosphine ligand. A catalyst analogous to the Pd-PE-g-PAA powder catalyst on polystyrene (Pd-PS-g-PAA) was similarly active for allylic substitution and could also be used in Heck reactions at 80-100°C in N,N-dimethylacetamide (DMA). Analysis of the product solutions for Pd leachate and a correlation of the Pd leaching with product formation in the allylic substitution chemistry for both types of catalysts suggests that the active catalysts in these reactions are leached from the support. In the case of the allylic substitution reaction, external triphenylphosphine and substrate together are required for the chemistry and Pd leaching.

Full text of DOI:10.1139/V06-076

1343
Catalysis with palladium colloids supported in
poly(acrylic acid)-grafted polyethylene and
polystyrene1
David E. Bergbreiter, Andrew Kippenberger, and Zhenqi Zhong
Abstract: Grafts of poly(acrylic acid) on polyethylene powder (PE-g-PAA) or polystyrene (PS-g-PAA) can be used to
support Pd(0) crystallites that function like a homogeneous Pd(0) catalyst in some reactions. These Pd–PE-g-PAA cata-
lysts were active in allylic substitution reactions in the presence of added phosphine ligand. A catalyst analogous to the
Pd–PE-g-PAA powder catalyst on polystyrene (Pd–PS-g-PAA) was similarly active for allylic substitution and could
also be used in Heck reactions at 80–100 °C in N,N-dimethylacetamide (DMA). Analysis of the product solutions for
Pd leachate and a correlation of the Pd leaching with product formation in the allylic substitution chemistry for both
types of catalysts suggests that the active catalysts in these reactions are leached from the support. In the case of the
allylic substitution reaction, external triphenylphosphine and substrate together are required for the chemistry and Pd
leaching.
Key words: catalysis, palladium, allylic substitution, grafted polystyrene, supported catalysts.
Résumé : Les produits de greffage d’acide polyacrylique sur de la poudre de polyéthylène (PE-g-APA) ou de polysty-
rène (PS-g-APA) peuvent être utilizés comme support de cristallites de Pd(0) qui, dans certaines réactions, peuvent
fonctionner comme catalyseur de Pd(0). Ces catalyseurs de Pd–PE-g-APA mis en présence d’un ligand phosphine sont
actifs dans les réactions de substitution allylique. Un catalyseur analogue au catalyseur en poudre Pd–PE-g-APA sur du
polystyrène (Pd–PS-g-APA) est aussi actif pour la réaction de substitution allylique et il peut aussi être utilizé dans les
réactions de Heck, à des températures allant de 80 à 100 °C, dans le N,N-diméthylacétamide (DMA). Une analyse des
solutions de produits pour rechercher la présence de Pd qui aurait été lessivé ainsi qu’une corrélation du Pd de lessi-
vage avec la formation de produit dans la chimie de la substitution allylique pour les deux types de catalyseur suggère
que les catalyseurs actifs dans ces réactions sont lessivés de leur support. Dans le cas de la réaction de substitution al-
lylique, la présence simultanée de la triphénylphosphine externe et du substrat sont requis pour la chimie et le lessivage
du Pd.
Mots clés : catalyse, palladium, substitution allylique, polystyrène greffé, catalyseurs sur un support.
[Traduit par la Rédaction] Bergbreiter et al. 1350
Introduction
and cross-coupling chemistry, are of more current interest. A
variety of palladium catalysts have been developed for these
purposes (2). Some years ago we had noted that classical
Pd–C catalysts, under relatively mild conditions (i.e., at
reaction temperatures <100 °C) (3, 4), exhibit reactivity re-
sembling that of the well-established homogeneous Pd cata-
lyst, tetrakis(triphenylphosphine)palladium(0). Specifically, we
found that both a supported Pd(0) species prepared from an
organometallic derivative of cross-linked polystyrene (5)
and simple Pd–C could effectively catalyze allylic substitu-
tion of allyl acetates by nucleophiles like secondary amines.
In these cases, catalysis was most effective in the presence
of a soluble phosphine with reactions sometimes occurring
at room temperature.
The importance of new sorts of catalysts, including adven-
titiously formed or designed colloidal metal catalysts for
various catalytic reactions including cross-coupling reac-
tions, has recently received increased attention (6–11). For
example, we and others have noted the exceptional reactivity
of Pd(0) colloidal catalysts, presumably generated in situ
from Pd(II) palladacycle compounds during cross-coupling
chemistry (12–17). These studies suggest that further explo-
The use of supported heterogeneous catalysts is well-
established. Palladium catalysts are among the most com-
mon examples of such species. For example, catalysts like
Pd–C are widely used for hydrogenation of alkenes (1).
They have the virtue of wide availability, relatively low costs
for handling and preparation, and activity that is often as
good or better than that of the more “novel” hydrogenation
catalysts reported from time to time.
While hydrogenation is an important reaction, other
palladium-catalyzed processes, such as allylic substitutions
Received 31 October 2005. Accepted 28 February 2006.
Published on the NRC Research Press Web site at
http://canjchem.nrc.ca on 12 August 2006.
D.E. Bergbreiter,2 A. Kippenberger, and Z. Zhong.
Department of Chemistry, P.O. Box 30012, College Station,
TX 77842-3012, USA.
1This article is part of a Special Issue dedicated to Dr. Alfred
Bader.
2Corresponding author (e-mail: bergbreiter@tamu.edu).
Can. J. Chem. 84: 1343–1350 (2006)
doi:10.1139/V06-076
© 2006 NRC Canada

1344
Can. J. Chem. Vol. 84, 2006
ration of ways to support and use Pd colloids merit attention.
Our studies of these putative Pd colloids formed from SCS–
Pd(II) precursors noted that those species were more soluble
in polar phases than in nonpolar phases (14). Since we re-
cently developed some new sorts of immobilized polar
poly(acrylic acid) graft phases on polyethylene powders and
on DVB cross-linked polystyrene resins, we have sought to
explore these supported phases as supports for catalytically
active Pd colloids. We have shown that Pd colloids formed
and supported in these poly(acrylic acid) phases are active
catalysts. While they are competent catalysts for hydro-
genations, their activity is not significantly different then the
more readily available and simpler Pd–C catalysts in hydro-
genation catalysis.
While catalytic activity, especially reactivity in a reaction
as simple as alkene hydrogenation, is often easy to demon-
strate and assay, the identity of the actual species that leads
to catalysis is not always clear. Indeed, it can even be diffi-
cult to determine whether the actual catalyst is soluble or in-
soluble (18). Such issues are even more complex when the
chemistry is carried out with supported palladium colloids.
For example, while we assumed a soluble Pd species was
generated in our earlier studies (3, 4), a three-phase test us-
ing a polymer-supported Pd species (a soluble phosphine)
and a second polymer-supported substrate failed to detect a
soluble catalyst. We rationalized this result at the time on the
basis that the actual phosphine-ligated Pd colloidal catalyst
could still be present but could be unreactive to a second in-
soluble supported substrate (3). More recent work from sev-
eral laboratories using various procedures, including Rebek’s
three-phase test (19), has been more successful in showing
that homogeneous catalysts are responsible for the catalysis
seen in other systems with “heterogeneous” catalysts (12,
19–24).
DVB cross-linked polystyrene beads (PS-g-PAA) as sup-
ports (19, 20, 25). Such colloidal catalysts have precedent
both in our earlier work with polystyrene-supported palla-
dium colloids (3, 4) and in other reports where supported
thin films or inorganic matrices support analogs of homoge-
neous catalysts (3, 4, 8, 26–31).
The synthesis and characterization of both the PE-g-PAA
and PS-g-PAA supports and catalysts were previously de-
scribed, and the syntheses of these catalysts are summarized
in Schemes 1 and 2. As shown in both schemes, the forma-
tion of a PE-g-PAA or a PS-g-PAA supported Pd colloid in-
volved ion exchange of palladium acetate onto a PE- or PS-
bound carboxylate group with subsequent reduction of the
Pd(II) carboxylate ionically attached to the polymer graft.
The Pd–PE-g-PAA and Pd–PS-g-PAA catalysts contained
about 0.01 or 1.1 mmol of Pd per gram of support, respec-
tively, based on digestion of the polymers and analysis of the
residue for Pd by ICP metal analysis. In the case of the Pd–
PE-g-PAA catalyst, imaging of the catalyst by XPS spectros-
copy showed that the Pd colloids were distributed through-
out a 50–100 µm region on the surface of the powder
(Fig. 1). In the case of the Pd–PS-g-PAA, the Pd colloidal
particles were mostly within the polystyrene bead.
The use of the Pd–PE-g-PAA and Pd–PS-g-PAA catalysts
for catalytic hydrogenation reactions were described previ-
ously (19, 20). Both catalysts were found to be fully recycla-
ble and no Pd leachate was detected in those studies. Of
more interest in this study was that these catalysts also cata-
lyze some Pd(0)-like reactions. Specifically, the PE-bound
Pd colloids were effective in promoting allylic substitution
reactions between secondary amines and allyl acetate
(eq. [1]). Such reactions were not successful if no added
phosphine was present in solution, but were successful in the
presence of 5 mol% added triphenylphosphine (Table 1).
Here we describe work where we have shown that the Pd
colloids in poly(acrylic acid) grafted polystyrene, which we
used previously in hydrogenation reactions, are active in
Heck catalysis and also in allylic substitution chemistry.
While the analogous polyethylene-supported catalysts can-
not be used in Heck chemistry (the required temperatures
are incompatible with the underlying polymer), the polyeth-
ylene-bound catalysts exhibit Pd(0)-like reactivity in allylic
substitutions of allylic acetates and amines. While we did
not achieve the sort of notable reactivity we and others have
seen in other putative Pd colloidal catalysts (12–17), we
have been able to show that, in the case of these Pd colloids
in a grafted polystyrene matrix or supported on polyethylene
powder, Pd leaching is associated with the observed Pd(0)-
like chemistry. The simplest explanation for these results is
that soluble phosphine-free (Heck catalysis) or phosphine-
ligated Pd species (allylic substitutions), derived from the
interfacially supported Pd colloids in both the poly(acrylic
acid)-grafted polyethylene powder and the poly(acrylic
acid)-grafted DVB cross-linked polystyrene resin, are the ac-
tive catalysts.
HNR'2
O
[1]
O
0.05 mol % Pd–PE-g-PAA
NR'2
R
PPh3,
55 °C
The reactivity in the presence of added phosphine ligands
of the Pd–PE-g-PAA catalyst could be due to activation of
the Pd colloids within the polymer matrix by the added
triphenylphosphine. Alternatively, formation of tetra-
kis(triphenylphosphine)palladium or formation of some less-
defined soluble phosphine-ligated Pd(0) species could
explain the observed chemistry. As noted above, we had pre-
viously been unsuccessful at using a three-phase test to de-
tect such soluble intermediates in a related system (3). Here
we used several other approaches that together provide con-
vincing evidence that a soluble Pd species is formed in these
catalytic reactions.
To ascertain the origin of the catalysis, several experi-
ments were performed. First, XPS analysis of Pd crystallites
in a hyperbranched graft of Pd–PE-g-PAA showed a Pd 3d3/2
peak at 333.3 eV. This peak was discernibly different than
the peak for a previously described immobilized molecular
catalyst — a Pd(0) complex prepared using the same poly-
mer that had been further modified to contain aminodi-
phenylphosphinopropyl ligands (i.e., Pd(0)-DPPA–PE-g-
PAA; DPPA is diphenylphosphinopropylamide containing a
phosphine ligand that was covalently coupled to poly(acrylic
Results and discussion
We recently described several sorts of supported palla-
dium catalysts using a poly(acrylic acid) graft on polyethyl-
ene powder (PE-g-PAA) or a poly(acrylic acid) graft within
© 2006 NRC Canada

Bergbreiter et al.
1345
Scheme 1. Synthesis of Pd–PE-g-PAA catalysts.
O
O
O
OH
O
O
O
PE
Powder
OEt
O
OtBu
CrO3
RNH2
100
ClCO2Et
OEt
H2NR
H2SO4
O
O
O
OH
O
O
O
O
O
O
NH
O
=
=
100
PTBA
OH
O
O
OH
100
PAA
O
CH3SO3H
O
O
NH
repeat
NH
O
O
O
NH2R-PTBA-RNH2
grafting, hydrolysis
O
and again repeat
OH
NH
NH2R-PTBA-RNH2
grafting, hydrolysis
O
O
O
Pd
NH
1. NaOH
Pd
O
O
2. Pd(OAc)2
3. H2
Pd
NH
O
Pd
Pd
Pd/PE-g-PAA
acid)-grafted PE via an amide bond) (25). No evidence was
obtained for the presence of phosphine in the recovered PE
support after a reaction, suggesting that added phosphine
was not sorbed into the resin. Catalysts analogous to the Pd–
PE-g-PAA could also be prepared by deliberate oxidation of
the phosphine ligands in a Pd(0)-DPPA–PE-g-PAA catalyst —
a process that converted the light yellow Pd(0)-DPPA–PE-g-
PAA powder into a grey powder and produced a new peak
for oxidized phosphine in XPS spectroscopy and a Pd(0)
species with a Pd 3d3/2 peak such as that seen for the Pd
crystallites in Pd–PE-g-PAA.
While the original Pd(0)-DPPA–PE-g-PAA catalyst was
active in allylic substitution, after oxidation, both this recov-
ered catalyst, now containing Pd(0) colloids, and the Pd–PE-
g-PAA catalyst behaved similarly in that they had no activity
in allylic substitution unless external phosphine was present
(Fig. 2). These reactions catalyzed by Pd(0) colloids were
also accompanied by Pd leaching. Using ICP–MS (induc-
tively coupled plasma – mass spectroscopy) analysis, ca. 2%
to 3% Pd leaching was seen in the reaction between allyl
acetate and diethylamine with external phosphine. Pd leach-
ing was not seen with Pd(0)-DPPA–PE-g-PAA catalysts that
© 2006 NRC Canada

1346
Can. J. Chem. Vol. 84, 2006
Scheme 2. Synthesis of Pd–PAA-g-PS catalysts.
H
H
Cl
Li
BuLi
PS
PS
H
acid
H
OC(CH3)3
PS
PS
O
Li
CO2C(CH3)3
H
H
Pd(OAc)2
PS
PS
CO2H
CO2H
PS-g-PAA
PS
CO2PdOAc
H2
. Pd(0)
CO2H
Pd/PS-g-PAA
Fig. 1. XPS image of a Pd–PE-g-PAA powder. The white regions
represent the presence of Pd 3d electrons. The left side of the
image is focused on a Pd–PE-g-PAA particle, and the edge of
the three-dimensional particle roughly transects the photograph.
Table 1. Allylic substitution reactions of allyl acetate with
Pd–PE-g-PAA.
PPh3 (mol%)
Time (h)
Conversion (%)a
′′
-R
b
Diethylamine
Diethylamine
Pyrrolidine
0.00
0.05
0.05
24
6/6
20
—
99, 99c
99
Note: Reactions were carried out at 55 °C on a 20 mmol scale, typi-
cally without any added solvent.
aGC analysis showed that the conversion of starting material to product
was 99%. Isolated yields of allylic amines in similar reactions were typi-
cally ca. 90%.
bNo product was detected.
cConversions for a second cycle using the Pd–PE-g-PAA catalyst with a
fresh addition of 5 mol% triphenylphosphine.
These PE-bound Pd(0) species could not be used in Heck
catalysis. No Heck catalysis was seen at 55 °C, and heating
to higher temperatures led to thermal reorganization of the
grafted PE powder (melting that produced a material that be-
came inactive in hydrogenation or allylic substitution with or
without added phosphine). However, immobilized
poly(acrylic acid) grafts on a more thermally robust poly-
mer, which are analogous to the Pd–PE-g-PAA powder cata-
lysts described previously, can be prepared on DVB cross-
linked polystyrene supports as shown in Scheme 2. Ion ex-
change of Pd(OAc)2 with the immobilized poly(acrylic acid)
followed by reduction, produced immobilized Pd(0) colloids
that had hydrogenation activity analogous to that seen with
Pd–C or to Pd crystallites immobilized within hyper-
branched poly(acrylic acid) grafts on PE powder (20).
While the PE-supported Pd colloids could not be used in
Heck catalysis because of temperature limitations, Pd
colloids supported on more thermally robust DVB cross-
linked polystyrene were active in Heck catalysis (eq. [2]).
contained covalently bound phosphine ligands within the
functional polymer support that formed analogs of a molecu-
lar Pd(0) catalyst (25). Taken together, these results are con-
sistent with an explanation such as that proposed previously
(3) for the activity of these Pd–PE-g-PAA colloidal cata-
lysts, which is based on leaching of a phosphine-ligated Pd
catalyst or a Pd cluster.
© 2006 NRC Canada

Bergbreiter et al.
Table 2. Heck catalysis with Pd–PS-g-PAA.
1347
Substrate aryl iodide
Heck acceptor
Catalyst
Yield (%)a
4-Iodotoluene
4-Iodotoluene
tert-Butyl acrylate
tert-Butyl acrylate
Pd–PS-g-PAA
Pd–Cb
99
75
Iodobenzene
4-Iodoanisole
4-Iodoanisole
4-Iodoanisole
Iodobenzene
4-Bromoacetophenone
4-Bromoacetophenone
tert-Butylstyrene
tert-Butylbenzene
N,N-Dimethylacrylamide
N,N-Dimethylacrylamide
tert-Butyl acrylate
Pd–PS-g-PAA
Pd–PS-g-PAA
Pd–PS-g-PAA
Pd–C
Pd–PS-g-PAA
Pd–PS-g-PAA
Pd–PS-g-PAA
100
94
95
36
100
81
tert-Butyl acrylate
N,N-Dimethylacrylamide
66
Note: Reactions were carried out on a 2 mmol scale in 10 mL of N,N-dimethylacrylamide using 2.5 mmol of Et3N
as a base at 80 °C for 36 h (or at 100 °C for 24 h). The Heck acceptor was present in 20 mol% excess. The PS-g-Pd
catalyst contained 0.11 mol% of Pd. The Pd–C species used was 10% Pd–C (Sigma-Aldrich). Yields were measured
by GC using dodecane as an internal standard. The catalyst was present at 0.1 mol%.
aIsolated yields were typically 10% lower for reactions carried out on a 5 or 10 mmol scale.
bThe catalyst was present at 0.5 mol%.
Fig. 2. Allylic substitution of allyl acetate using diethylamine
with a Pd(0) catalyst containing Pd(0) crystallites and no phos-
phine ligands within the interface. After 24 h, external PPh3 was
introduced to the reaction mixture, a step that did lead to allylic
substitution product formation.
Table 3. Recycling experiments using Pd–PS-g-PAA in the Heck
reaction of tert-butyl acrylate and iodobenzene (eq. [2]).
Cycle
Conversion (%)
Time (h)
1
2
3
4
99
95
46
99
6
6
6
9
100
80
60
40
20
0
Note: After the third cycle, the catalyst was recovered and washed in a
Soxhlet apparatus with MeOH for 24 h.
tion after centrifugation. Fresh reagents and solvent were
then added to the catalyst for the second cycle (Table 3). Af-
ter 6 h, the reaction had proceeded to 95% conversion by
GC analysis. During the third recycling experiment the reac-
tion went to only 46% conversion after 6 h. After washing
the catalyst by Soxhlet extraction with MeOH for 24 h, a
fourth cycle went to 99% conversion after 9 h. Presumably,
the reaction salts had built up in the resin, inhibiting the re-
action. Such fouling of a recovered catalyst by salt by-
products is a potential general problem for any recoverable
catalyst that employs a relatively polar phase.
Trace metal analysis (ICP–MS) of an acid-digested
aliquot from the first reaction cycle showed 2.1% leaching
of Pd. This result is in accord with other suggestions that ac-
tivity of a heterogeneous Pd catalyst involves oxidative addi-
tion and formation of a soluble Pd species (2).
The Pd–PS-g-PAA catalyst is also active for allylic substi-
tution chemistry (Table 4). As was true for Pd–PE-g-PAA
catalysts, the presence of external triphenylphosphine is re-
quired for these reactions to proceed. A series of allylic sub-
stitution reactions using allyl acetate or cinnamyl acetate
reacting with diethylamine or piperidine were performed
(eq. [3]). The reaction between allyl acetate and diethyl-
amine went to only 38% conversion in 18 h without the
presence of external triphenylphosphine. With external phos-
phine, the reaction goes to complete conversion in 1 h. The
reaction between cinnamyl acetate and diethylamine was
even more dependant on external phosphine. After 24 h
without any external phosphine, no reaction was seen by GC
analysis. A reaction in the presence of added triphenyl-
phosphine was completed within 8 h.
0
10
20
30
40
Time (h)
As shown in Table 2, good yields in small scale reactions
were obtained with a variety of aryl iodides. However,
unactivated aryl bromides and aryl chlorides were not sub-
strates, even if added to a reaction along with a more reac-
tive aryl iodide. The activity of the Pd–PS-g-PAA catalysts
was also compared to Pd–C — a simple catalyst that also
has modest activity in Heck catalysis. The PS-g-PAA immo-
bilized Pd colloids were found to be more active. These
cross-coupling reactions occurred without any added phos-
phine ligand present.
0.1 mol %
I
Pd–PS-g-PAA
Et3N
O
[2]
O
+
DMA, 100 °C
O
O
Recycling the Pd species in a polystyrene resin in Heck
catalysis was tested using the Heck reaction between
iodobenzene and tert-butyl acrylate (eq. [2]) as a model re-
action. The first cycle was complete in 6 h by GC analysis.
The catalyst was recovered by pouring off the reaction solu-
© 2006 NRC Canada

1348
Can. J. Chem. Vol. 84, 2006
Table 4. Results for allylic substitution reactions with Pd–PS-g-PAA.
-R
-R′
PPh3 (mmol)
Time (h)
Conversion (%)
-H
-CH2CH3
—
18
1
38
99
0
-H
-CH2CH3
0.1
—
-Ph
-Ph
-H
-CH2CH3
24
8
-CH2CH3
0.1
0.05
99
99
-CH2CH2CH2CH2CH2-
1
Note: These reactions were carried out in THF at 60 °C on a 10 or 20 mmol scale.
HNR'2
O
Fig. 3. A Pd–PS-g-PAA catalyzed allylic substitution reaction
between cinnamyl acetate and diethyl amine carried out in the
presence of 0.5 mol% of added triphenylphosphine in THF at
60 °C. The various plots show the extent of reaction in the pres-
ence of added triphenylphosphine as a function of reaction time
(᭹), the concentration of Pd in the leachate as a function of
time (᭺), and the concentration of leachate in a reaction with
the same catalyst and the same amount of triphenylphosphine
and amine but no cinnamyl acetate substrate (ٗ). No reaction
was observed in the absence of triphenylphosphine over a 24 h
period under the same reaction conditions.
[3]
NR'2
R
R
O
0.5 mol % Pd–PS-g-PAA
PPh3,
55 °C
A more detailed study showed that these reactions appear
to involve Pd leachates as the active catalyst. An induction
period of ca. 2 h was observed for the formation of the
cinnamyl acetate and diethylamine product. Interestingly, the
progress of these allylic substitution reactions using Pd–PS-
g-PAA catalysts with added triphenylphosphine were corre-
lated with leaching of palladium into solution. In these
studies, the first experiment was designed to monitor the Pd
leaching into solution from the Pd–PS-g-PAA catalyzed
allylic substitution reaction between cinnamyl acetate and
diethylamine with triphenylphosphine present. During the
course of the 8 h reaction between cinnamyl acetate and
diethylamine, aliquots were removed from solution. The
aliquots were digested in acid and analyzed by ICP–MS.
Figure 3 shows the results from this experiment. The results
show very little Pd leaching into solution during the induc-
tion period. As product forms, the Pd concentration in the
reaction mixture increases. A second control reaction that
had only triphenylphosphine and Pd–PS-g-PAA catalyst
showed almost no palladium leaching for 8 h, the time re-
quired for complete conversion of starting materials to prod-
uct. This suggests that the substrates cause the leaching of
Pd and that the active catalytic species is generated with the
presence of triphenylphosphine. In this case the Pd–PS-g-
PAA catalyst serves as a reservoir of Pd for the reaction.
These results are in accord with previous results from our
group where the three-phase test for a soluble Pd catalyst
failed using a polymer-supported allylic ester substrate and
Pd–C- and polystyrene-supported Pd(0) catalysts (3). These
results showed no reaction between a polystyrene-bound
allyl ester and diethylamine using a heterogeneous Pd cata-
lyst with triphenylphosphine in solution. However, as shown
here, that result is understandable since in that experiment
the substrate and Pd source were both heterogeneous. As-
suming that this is also the case presently, the soluble phos-
phine alone would not cause the Pd leaching and no reaction
would occur.
100
80
60
40
20
0
16
14
12
10
8
6
4
2
0
0
2
4
6
8
Time (h)
phine ligand. A catalyst analogous to the Pd–PE-g-PAA
powder catalyst on polystyrene (Pd–PS-g-PAA) was simi-
larly active for allylic substitution and could also be used in
Heck reactions at 80 °C in N,N-dimethylacetamide (DMA).
Analysis of the solution for Pd and a correlation of the Pd
leaching with product formation in the allylic substitution
chemistry for both types of catalysts suggests that the active
catalysts in these reactions are leached from the support. In
the case of the allylic substitution reaction, external
triphenylphosphine and substrate together are required for
the chemistry and Pd leaching.
Experimental section
General methods
Merrifield Resin (RGEN100) was obtained from the
American Peptide Company, Inc. (Sunnyvale, California) as
a 100–200 mesh resin and had a loading of 1.05 mmol of
ClCH2 groups per gram. Other reagents and solvents were
obtained from the Sigma-Aldrich Chemical Company and
were used as received unless noted otherwise. X-ray photo-
electron spectra were obtained on a Kratos Axis Ultra XPS
Conclusion
Hyperbranched grafts of poly(acrylic acid) on polyethyl-
ene powder or polystyrene can be used to support Pd(0)
crystallites that function as a homogeneous Pd(0) catalyst in
some reactions. These Pd–PE-g-PAA catalysts were active in
allylic substitution reactions in the presence of added phos-
© 2006 NRC Canada

Bergbreiter et al.
1349
(Manchester, UK) using a monochromatic Al Kα source
(400 W) in a UHV environment (ca. 5 × 10–9 torr (1 torr =
133.3224 Pa). During acquisition, surfaces were kept from
charging by the application of low energy electrons. Surface
elemental composition was determined by normalized inte-
gration of the resulting peaks using Kratos software (Kratos
and agitated with a wrist-action shaker. The progress of the
reaction was checked using GC.
General procedure for Heck catalysis, recycling, and
leachate analysis using Pd–PS-g-PAA
Iodobenzene (5 mmol), tert-butyl acrylate (6 mmol),
triethylamine (7.5 mmol), and 0.5 mol% Pd(0)-PAA–PS
(0.025 mmol of Pd, 22.5 mg of resin) were added to a
40 mL centrifuge tube along with 25 mL of DMA. This mix-
ture was heated to 100 °C and agitated using a wrist-action
shaker until GC analysis showed the reaction was complete.
After the reaction was complete, the catalyst was separated
by centrifugation and decantation. The catalyst isolated in
this manner could then be reused in a subsequent reaction.
To analyze for Pd leaching, ca. 25% of the decanted solution
from this reaction was evaporated to dryness and then heated
on a hot plate with 20 mL of concentrated sulfuric acid for
24 h. The solution was diluted to 50 mL using 1% HCl and
analyzed by ICP–MS. In a typical analysis, a reaction that
started with 2.65 mg of Pd had 56 µg of Pd in total in this
solution corresponding to 2.1% leaching.
1
Analytical, Chestnut Ridge, New York). H and 13C NMR
spectroscopy experiments were carried out using Mercury
300, Unity 300, or Unity 500 spectrometers. GC analyses
were carried out using a Shimazdu Model 2010 GC. All
glassware used in the experiments in which trace metal anal-
ysis was performed was soaked in a 1 mol/L HNO3 acid
bath for at least 12 h prior to the experiment.
Preparation of Pd–PS-g-PAA
A reported procedure was used (20). The product Pd cata-
lyst was prepared on a 12 g scale. The Pd loading of the
Pd(II)–PS-g-PAA prepared in this manner was 1.27 mmol/g,
based on gravimetric analysis of the exchange of Pd(OAc)2
and the sodium salt of the poly(acrylic acid)-grafted polysty-
rene. Subsequent reduction of this Pd(II) salt in the
poly(acrylic acid) graft produces a product (Pd–PS-g-PAA)
that was used in the catalytic studies. Analysis by ICP–MS
of a solution of the residue after combustion and acid diges-
tion of this Pd–PS-g-PAA product in triplicate showed that
the Pd loading of this polystyrene-supported Pd product was
1.1 mmol of Pd per gram of Pd–PS-g-PAA.
General procedure for allylic substitution reactions and
leachate analysis using Pd–PS-g-PAAC
Cinnamyl acetate (10 mmol, 1.76 g) and diethylamine
(24 mmol, 1.75 g) were combined with triphenylphosphine
(0.2 mmol, 52.4 mg) and Pd(0)-PAA–PS (0.5 mol%, 45 mg
of resin) in 20 mL of distilled THF in a 40 mL centrifuge
tube. This mixture was degassed by freeze–pump–thaw cy-
cles three times and then backfilled with N2. The reaction
tube was heated to 55 °C and agitated using a wrist-action
shaker. A control reaction was also carried out at the same
time using the same amount of catalyst, phosphine, and sol-
vent but without any cinnamyl acetate. GC analysis of
aliquots was carried out as the reaction progressed. The re-
mainder of each aliquot was digested by heating in concd.
H2SO4 for 24 h. The concentrated digest was diluted to
25 mL with 1% HCl and then analyzed by ICP–MS.
Preparation of Pd–PE-g-PAA powder
Following a literature procedure (19), a poly(acrylic acid)
hyperbranched graft on a PE powder sample (1.3 g) was first
stirred with a pH 9 buffer for 2 h to deprotonate the -CO2H
groups of the graft. The resulting carboxylate-containing
polymer was then suspended in 25 mL of an acetone–water
(4:1) mixture with 20 mg of Pd(OAc)2. This solution was
stirred at room temperature for 24 h then filtered and washed
with dilute acidic ethanol and ethanol. The yellowish powder
was suspended in EtOH and reduced with H2 in a Parr appa-
ratus to yield the Pd–PE-g-PAA catalyst that was used in
these studies. To determine the Pd loading, this powder sam-
ple was combusted in a crucible at 650 °C for 2 h. The resi-
due was digested in acid and ICP–MS analysis of the residue
for Pd showed that Pd loading was ca. 0.01 mmol of Pd per
gram of Pd–PE-g-PAA powder.
Pd–PE-g-PAA catalyzed allylic substitution reactions
Allyl acetate (18.8 mmol, 1.8852 g), piperidine
(37.8 mmol, 3.2125 g), triphenylphosphine (0.01 mmol,
26.4 mg), and 400 mg of Pd–PE-g-PAA powder were com-
bined in a 20 mL flask. This mixture was degassed three
times by freeze–pump–thaw, backfilled with N2, and then
heated to 55 °C. After 20 h, the allyl acetate had been com-
pletely consumed based on GC analysis.
Typical procedure for Heck catalysis with 10% Pd–C
or Pd–PS-g-PAA
An aryl iodide such as iodotoluene (2.0 mmol), 2.4 mmol
(0.3139 g, 98%) of tert-butyl acrylate, 2.5 mmol of
triethylamine, along with 0.01 mmol of the Pd catalyst (e.g.,
9 mg of Pd–PS-g-PAA) were added to a 50 mL flask. Then
10 mL of DMA was added to the mixture and the reaction
was heated to 80 °C.
Acknowledgments
We thank the Robert A. Welch Foundation (A-0639) and
the National Science Foundation (CHE-0446107) for support
of this research.
Typical procedure for allylic substitutions with Pd–PS-
g-PAA
References
A mixture of 2.0 mmol (0.36 g) of cinnamyl acetate,
2.4 mmol (0.21 g) of morpholine, 0.1 mmol (0.264 g) of
triphenylphosphine, and 0.01 mmol (9.0 mg) of Pd–PS-g-
PAA were added to a 50 mL reaction flask along with
10 mL of THF. The reaction mixture was heated to 60 °C
1. R.L. Augustine. Heterogeneous catalysis for the synthetic
chemist. Dekker, New York. 1995.
2. V. Farina. Adv. Synth. Catal. 346, 1553 (2004).
3. D.E. Bergbreiter, B. Chen, and D. Weatherford. J. Mol. Catal.
74, 409 (1992).
© 2006 NRC Canada

1350
Can. J. Chem. Vol. 84, 2006
4. D.E. Bergbreiter and B. Chen. J. Chem. Soc. Chem. Commun.
1238 (1983).
18. D.E. Bergbreiter, J.G. Franchina, and K. Kabza. Macro-
molecules, 32, 4993 (1999).
5. D.E. Bergbreiter, B. Chen, and T.J. Lynch. J. Org. Chem. 48,
4179 (1983).
19. D.E. Bergbreiter, G. Tao, and A.M. Kippenberger. Org. Lett. 2,
2853 (2000).
6. M.T. Reetz and J.G. de Vries. Chem. Commun. (Cambridge),
1559 (2004).
20. D.E. Bergbreiter and Z. Zhong. Ind. Eng. Chem. Res. 44, 8616
(2005).
7. B.H. Lipshutz. Adv. Synth. Catal. 343, 313 (2001).
8. C.K.Y. Lee, A.B. Holmes, S.V. Ley, I.F. McConvey, B. Al-
Duri, G.A. Leeke, R.C.D. Santos, and J.P.K. Seville. Chem.
Commun. (Cambridge, U.K.), 2175 (2005).
9. S.J. Broadwater, S.L. Roth, K.E. Price, M. Kobaslija, and D.T.
McQuade. Org. Biomol. Chem. 3, 2899 (2005).
10. P.D. Stevens, J. Fan, H.M.R. Gardimalla, M. Yen, and Y. Gao.
Org. Lett. 7, 2085 (2005).
11. F. Wen, H. Bonneman, J.Y. Jiang, D.M. Lu, Y.H. Wang, and
Z.L. Jin. Appl. Organomet. Chem. 19, 81 (2005).
12. I.P. Beletskaya and A.V. Cheprakov. J. Organomet. Chem. 689,
4055 (2004).
21. C.M. Hagen, J.A. Widegren, P.M. Maitlis, and R.G. Finke. J.
Am. Chem. Soc. 127, 4423 (2005).
22. J. Rebek, Jr. Tetrahedron, 35, 723 (1979).
23. I.W. Davies, L. Matty, D.L. Hughes, and P.J. Reider. J. Am.
Chem. Soc. 123, 10139 (2001).
24. A. Corma, D. Das, H. Garcia, and A. Leyva. J. Catal. 229, 322
(2005).
25. D.E. Bergbreiter, A.M. Kippenberger, and G. Tao. Chem.
Commun. (Cambridge), 2158 (2002).
26. C.M. Crudden, M. Sateesh, and R. Lewis. J. Am. Chem. Soc.
127, 10045 (2005).
27. J.P. Arhancet, M.E. Davis, J.S. Merola, and B.E. Hanson. J.
13. K. Yu, W. Sommer, J.M. Richardson, M. Weck, and C.W.
Jones. Adv. Synth. Catal. 347, 161 (2005).
14. D.E. Bergbreiter, P. Osburn, and J.D. Frels. Adv. Synth. Catal.
347, 172 (2005).
Catal. 121, 327 (1990).
28. F. Quignard, A. Choplin, and A. Domard. Langmuir, 16, 9106
(2000).
29. S.-I. Fujita, Y. Sano, M.B. Bhanage, and M. Arai. J. Catal.
15. W. Sommer, K. Yu, J.S. Sears, Y. Ji, X. Zhang, R.J. Davis,
D.C. Sherrill, C.W. Jones, and M. Weck. Organometallics, 24,
4351 (2005).
16. D. Olsson, P. Nilsson, M. El Masnaouy, and F.O. Wendt. Dal-
ton Trans. 1924 (2005).
225, 95 (2004).
30. Z.M. Michalska, L. Rogalski, K. Rozga-Wijas, J. Chojnowski,
W. Fortuniak, and M. Scibiorek. J. Mol. Catal. A: Chem. 208,
187 (2004).
31. C.P. Mehnert. Chem. Eur. J. 11, 50 (2005).
17. C.S. Consorti, F.R. Flores, and J. Dupont. J. Am. Chem. Soc.
127, 12054 (2005).
© 2006 NRC Canada