Pd catalysis on dendronized solid support: Generation effects and the influence of the backbone structure

Article abstract of DOI:10.1021/ja065265d

Recent studies revealed that catalysts, prepared on dendronized support, frequently exhibit enhanced activity and selectivity as compared to their non-dendronized analogues. Regretfully, in early studies of the supported dendritic catalysis, no particular attention was paid to the coordinative nature of the dendritic backbone. In this study, we functionalized Wang polystyrene support with three types of dendritic templates: poly(aril benzyl ether), poly(aryl benzyl thioether), and poly(aryl benzyl amine). These dendronized resins were further decorated with phosphine ligands on the periphery and complexed with a Pd(0) catalytic precursor. The catalysis of the Heck and Suzuki reactions of bromobenzene with the first to third generation supported dendritic catalysts was examined and compared to that of the non-dendritic analogues. All of the examined reactions revealed a positive dendritic effect, reflected in up to 5-fold increase in yield, in the most prominent case. The reasons for the observed effect are the proximity of the ligating sites translated into reduced cross-linking and, probably, the increased distance of the catalyst from the polymer matrix. We proved, however, that the latter could not be achieved with a linear spacer. Although the Suzuki reaction was rather insensitive to the backbone structure, the Heck reaction catalysis at 80°C exhibited substantial sensitivity to the nature of the dendritic backbone, with the polyether structure demonstrating the best outcome. This is the first demonstration of the influence of the coordinative ability of the backbone on the activity of a supported dendritic catalyst.

Full text of DOI:10.1021/ja065265d

Published on Web 04/18/2007
Pd Catalysis on Dendronized Solid Support: Generation
Effects and the Influence of the Backbone Structure
Adi Dahan and Moshe Portnoy*
Contribution from the School of Chemistry, Raymond and BeVerly Sackler Faculty of Exact
Sciences, Tel AViV UniVersity, Tel AViV 69978, Israel
Received July 23, 2006; E-mail: portnoy@post.tau.ac.il
Abstract: Recent studies revealed that catalysts, prepared on dendronized support, frequently exhibit
enhanced activity and selectivity as compared to their non-dendronized analogues. Regretfully, in early
studies of the supported dendritic catalysis, no particular attention was paid to the coordinative nature of
the dendritic backbone. In this study, we functionalized Wang polystyrene support with three types of dendritic
templates: poly(aril benzyl ether), poly(aryl benzyl thioether), and poly(aryl benzyl amine). These
dendronized resins were further decorated with phosphine ligands on the periphery and complexed with a
Pd(0) catalytic precursor. The catalysis of the Heck and Suzuki reactions of bromobenzene with the first
to third generation supported dendritic catalysts was examined and compared to that of the non-dendritic
analogues. All of the examined reactions revealed a positive dendritic effect, reflected in up to 5-fold increase
in yield, in the most prominent case. The reasons for the observed effect are the proximity of the ligating
sites translated into reduced cross-linking and, probably, the increased distance of the catalyst from the
polymer matrix. We proved, however, that the latter could not be achieved with a linear spacer. Although
the Suzuki reaction was rather insensitive to the backbone structure, the Heck reaction catalysis at 80 °C
exhibited substantial sensitivity to the nature of the dendritic backbone, with the polyether structure
demonstrating the best outcome. This is the first demonstration of the influence of the coordinative ability
of the backbone on the activity of a supported dendritic catalyst.
Introduction
Repeatedly, it was demonstrated that the supporting matrix
crucially influences the performance of the catalyst. Recent
The quest for easily separable catalysts is driven by economic
considerations as well as by increasing environmental concerns.
Along with relatively new approaches to catalyst separation (e.g.,
fluorous-phase soluble catalysts, dendrimer-, or other soluble
polymer-supported catalysts),1 the traditional immobilization of
analogues of homogeneous catalysts on insoluble polymers has
received special attention.2 In recent years, research on polymer-
bound catalysts was assisted by the enormously expanded
volume of reliable procedures for solid-phase synthesis.3 Ad-
ditional motivation to develop such catalysts was provided by
the emerging technique of solution-phase combinatorial chem-
istry, based on insoluble reagents, catalysts, and scavengers.4
Unfortunately, in most cases, the immobilization of catalytic
species on a polymer support is accompanied by a significant
loss in catalytic activity and/or selectivity. Polymer-based
catalysts are often ineffective, due to leaching of the catalytic
metals, the unavailability and nonuniformity of the catalytic
sites, and indefinable incorporation of catalysts within the
polymer bulk.
studies revealed that catalysts, prepared on dendronized supports,
can exhibit enhanced activity, selectivity, and recyclability (not
necessarily together), relative to their analogues, attached to a
nondendronized matrix.5 In some cases, the improvement was
generation-dependent and defined as the positive dendritic effect.
Dendrimers are branched, highly ordered oligomers, as-
sembled in a modular fashion from polyfunctional building
blocks and utilized for a surprisingly wide spectrum of applica-
tions, including catalysis.6 While the majority of dendritic
structures are prepared using orthodox solution chemistry,
increasing attention has recently been diverted to dendron
preparation on solid support.7 Solid-phase synthesis provides
solutions to a number of problems associated with dendrimer
preparation, particularly the time-consuming purifications, and
can also markedly improve the yield and homogeneity of the
formed dendrons.
(5) (a) Dahan, A.; Portnoy, M. Chem. Commun. 2002, 2700. (b) Dahan, A.;
Portnoy, M. Org. Lett. 2003, 5, 1197. (c) Bourque, S. C.; Alper, H.; Manzer,
L. E.; Arya, P. J. Am. Chem. Soc. 2000, 122, 956. (d) Alper, H.; Arya, P.;
Bourque, S. C.; Jefferson, G. R.; Manzer, L. E. J. Am. Chem. Soc. 2001,
123, 2889. (e) Lu, S.-M.; Alper, H. J. Am. Chem. Soc. 2003, 125, 13126.
(f) Bu, J.; Judeh, Z. M. A.; Ching, C. B.; Kawi, S. Catal. Lett. 2003, 85,
183. (g) Chung, Y.-M.; Rhee, H.-K. Chem. Commun. 2002, 238.
(6) (a) Fre´chet, J. M. J.; Tomalia, D. A. Dendrimers and other dendritic
polymers; Wiley-VCH: New York, 2002. (b) Newkome, G. R.; Moore-
field, C. N.; Vo¨gtle, F. Dendrimers and dendrons: concepts, synthesis,
application; Wiley-VCH: Weinheim, 2001.
(1) See special issue of Chemical Reviews: Recoverable catalysts and reagents.
Gladysz, J. A., Guest Ed. Chem. ReV. 2002, 102.
(2) (a) McNamara, A. C.; Dixon, M. J.; Bradley, M. Chem. ReV. 2002, 102,
3275. (b) Leadbeater, N. E.; Marco, M. Chem. ReV. 2002, 102, 3217.
(3) Zaragoza-Do¨rwald, F. Organic Synthesis on Solid Phase - Supports,
Linkers, Reactions, 2nd ed.; Wiley-VCH: Weinheim, 2002.
(4) Ley, S. V.; Baxendale, I. R.; Bream, R. N.; Jackson, P. S.; Leach, A. G.;
Longbottom, D. A.; Nesi, M.; Scott, J. S.; Storer, R. I.; Taylor, S. J. J.
Chem. Soc., Perkin Trans. 2002, 1, 3815.
(7) Dahan, A.; Portnoy, M. J. Polym. Sci., Part A: Polym. Chem. 2005, 43,
235.
9
5860
J. AM. CHEM. SOC. 2007, 129, 5860-5869
10.1021/ja065265d CCC: $37.00 © 2007 American Chemical Society

Pd Catalysis on Dendronized Solid Support
A R T I C L E S
Scheme 1a
Most of the dendrons prepared on solid support are based on
carbonyl- and amine-containing moieties (e.g., polylysine
prepared by Tam, poly(amidoamine)(PAMAM) and poly-
(amidourea) prepared by Bradley).8 Although it is likely that
carbonyls and amines can coordinate to catalytic metal centers,
all early studies of supported dendritic catalysts were based on
polyamido backbones, particularly PAMAM dendrons.5c-g,9,10
Some of these works demonstrated the influence of the dendritic
backbone structure on the catalysis results,5c-e but no particular
attention was paid to the coordinating nature of the dendritic
backbone.
In our work, we decided to focus on polyether dendrons,
because ethers are sufficiently stable and demonstrate poor
ligating ability to transition metals and, particularly, late
transition metals, which are the focus of our catalytic studies.
We also decided to examine whether the coordinative nature
of the dendritic backbone influences the behavior of the dendritic
catalysts, and to generate dendrons that are based on thioether
and amine linkage between the modules, in addition to those
based on ether linkage. Herein, we report the study of the
supported complexes based on these three types of dendrons,
including the differences in their catalytic performance. Descrip-
tion of the synthesis and properties of the poly(aryl benzyl)-
ether dendrons has been published, as well as preliminary reports
on polythioether dendron synthesis and the Heck reaction with
polyether dendron-based catalysts.5b,11
a Reagents and conditions: (i) 1, 2, DCM, room temperature, 24 h; (ii)
LiBH4, B(OMe)3, THF, 67 °C, 12 h.
Scheme 2 a
Results and Discussion
The Poly(aryl benzyl)ether Dendron. The first synthetic
route to this dendron on polystyrene was reported by Bradley.12
However, it was based on a monomer that is not commercially
available and must be prepared in four steps. On the other hand,
as we communicated earlier, we were able to grow first to third
generation dendrons of this type (GnO) through a repetitive
Mitsunobu coupling-ester reduction sequence, while using the
commercially available dimethyl 5-hydroxyphthalate (2, Scheme
1).11a,13 Remarkably, use of the triphenylphosphine-sulfonamide
betain 1 and LiBH4/B(OMe)3 reducing agent in the two sequence
steps, respectively, leads to outstanding yields and purity.
The structure and purity of the dendrons were determined
a Reagents and conditions: (i) dimethylthiocarbamoyl chloride, DABCO,
DMA, 75 °C, 85%; (ii) 280 °C, 10 min, 95%.
able to convert the commercially available 5-hydroxy isoph-
thalate diester 2 into 4, the dimethylcarbamoyl derivative of 3,
via the Newman-Kwart rearrangement (Scheme 2).14 The
procedure is high-yielding, and a multigram scale preparation
of 4 is possible. Because of the strong propensity of 3 (formed
upon deprotective methanolysis of 4) to undergo oxidative
dimerization, we used 4 in the relevant dendron building
procedures, forming 3 in situ.
The dendrimer-forming synthetic sequence included three
repetitive steps: substitution of benzyl halide by the thiophe-
nolate, reduction, and chlorodehydroxylation (Scheme 3). The
last step reforms the benzyl halide terminal function, enabling
the construction of the next generation of the dendrimer, and
was cleanly accomplished with a PPh3/C2Cl6 mixture.
1
via TFA-induced cleavage, followed by H NMR, 13C NMR,
and MS analysis. According to the cleavage analysis and
quantification, the overall yield of the third generation is 82%.
The Poly(aryl benzyl)thioether Dendron. As we recently
communicated, we prepared the poly(aryl benzyl)thioether
dendrons GnS, the sulfur analogue of the polyether dendrons
GnO.11b The relevant monomer 3 is not commercially available,
and its synthesis has never been reported. Luckily, we were
(8) (a) Tam, J. P. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 5409. (b) Swali, V.;
Wells, N. J.; Langley, G. J.; Bradley, M. J. Org. Chem. 1997, 62, 4902.
(c) Fromont, C.; Bradley, M. Chem. Commun. 2000, 283.
(9) (a) Bourque, S. C.; Maltais, F.; Xiao, W.-J.; Tardif, O.; Alper, H.; Arya,
P.; Manzer, L. E. J. Am. Chem. Soc. 1999, 121, 3035. (b) Alper, H.; Arya,
P.; Bourque, S. C.; Jefferson, G. R.; Manzer, L. E. Can. J. Chem. 2000,
78, 920. (c) Reynhardt, J. P. K.; Alper, H. J. Org. Chem. 2003, 68, 8353.
(d) Chung, Y.-M.; Rhee, H.-K. C. R. Chim. 2003, 6, 695.
The synthesis was monitored, and the products characterized,
using gel-phase 13C NMR as well as acidolytic cleavage,
followed by 1H and 13C NMR, and mass spectrometry. Contrary
to the case of polyether dendrons, in all experiments, the
cleavage occurred between the resin and the Wang linker and
not between the linker and the dendron (Scheme 4a). Thus, a
4-hydroxybenzyl-protected version of the dendron was always
(10) Arya, P.; Rao, N. V.; Singkhonrat, J.; Alper, H.; Bourque, S. C.; Manzer,
L. E. J. Org. Chem. 2000, 65, 1881.
(11) (a) Dahan, A.; Portnoy, M. Macromolecules 2003, 36, 1034. (b) Dahan,
A.; Weissberg, A.; Portnoy, M. Chem. Commun. 2003, 1206.
(12) Basso, A.; Evans, B.; Pegg, N.; Bradley, M. Chem. Commun. 2001, 697.
(13) For use of Mitsunobu coupling in the synthesis of polyether dendrons in
solution, see: (a) Zeng, F.; Zimmerman, S. C. J. Am. Chem. Soc. 1996,
118, 5326. (b) Labbe, G.; Forier, B.; Dehaen, W. J. Chem. Soc., Chem.
Commun. 1996, 2143.
(14) (a) Newman, M. S.; Karnes, H. A. J. Org. Chem. 1966, 31, 3980. (b) Hioki,
H.; Still, W. C. J. Org. Chem. 1998, 63, 904.
9
J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007 5861

A R T I C L E S
Scheme 3a
Dahan and Portnoy
a Reagents and conditions: (i) 4, NaOMe 0.5 M in methanol, DMF, 70 °C, 12 h; (ii) LiBH4, B(OMe)3, THF, 67 °C, 12 h; (iii) C2Cl6, PPh3, THF, room
temperature, 6 h.
Scheme 4
signals of the isophthalate diester along with the appearance of
-CH2OH/-CH2OCOCF3 peaks (4.8 and 5.4 ppm, respectively).
The latter disappear completely upon chlorodehydroxylation and
are replaced by the -CH2Cl signals (4.5 ppm).
The Poly(aryl benzyl)amine Dendron. Following the syn-
thesis of the thioether analogue, we proceeded to prepare the
poly(aryl benzyl amine) dendrons. The synthetic procedure was
based on the substitution of the halide of the Wang Bromo resin
by the commercial dimethyl 5-aminoisophthalate (5) and
reduction of the ester groups with LiBH4 as the first two steps
(Scheme 5). Initially, we attempted to construct the dendron
by chlorodehydroxylation of the alcohols of G1N(OH), followed
by another nucleophilic substitution to yield the next generation
dendron. While the first of these two reactions gave a very clean
product, G1N(Cl), we found that the formed benzylic halides
cannot be effectively substituted by 5. Mild conditions left G1N-
(Cl) unchanged, while harsher conditions led to a mixture of
products. The poor nucleophilicity of the aniline unit is likely
to be responsible for the failure of this reaction.
released from the resin. While the linker is clearly visible in
1
the H and 13C NMR spectra of the cleavage solution, it is
removed under the CI-MS conditions, and disulfide-bridged
dendron dimers are observed in the MS spectra.
1
The aforementioned H NMR measurements demonstrated
excellent conversion and purity of each of the three repetitive
steps as per the following characteristic changes. Immobilization
According to the alternative strategy, the synthesis starts with
the halide substitution and reduction, as previously mentioned,
but the alcohols are then oxidized to aldehydes using sulfur
trioxide pyridine complex (Scheme 6). The next generation of
of 3 is accompanied by the complete disappearance of the
-CH2Cl signals (4.5 ppm) and appearance of the low-field
aromatic signals of the peripheral isophthalate units (8.5 and
8.2 ppm). Reduction results in complete disappearance of the
Scheme 5a
a Reagents and conditions: (i) 5, 2,6-lutidine, TBAI, DMF, 80 °C, 3 days; (ii) LiBH4, B(OMe)3, THF, 67 °C, 12 h; (iii) C2Cl6, PPh3, THF, room temperature,
6 h.
9
5862 J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007

Pd Catalysis on Dendronized Solid Support
A R T I C L E S
Scheme 6a
The hydroxy-terminated resin could not be analyzed using
gel-phase 13C NMR even for the first generation and even in
polar solvents, such as dioxane-d8. The density of the OH and
NH groups seems to generate a hydrogen-bonded hydrophilic
“film” that prevents the proper swelling of the polystyrene core
polymer. Significantly improved gel-phase spectra were ob-
served only after acylation of the OH and NH groups with acetyl
chloride.
For quantitative determination of the dendron loading, we
used two main strategies. First, we functionalized the hydroxy-
terminated resins with Fmoc-Gly-OH and photometrically
determined the Fmoc loading.15 According to the second
strategy, we esterified the hydroxy-terminated dendrons with
4-(diphenylphosphino)benzoic acid. Because we optimized this
reaction’s conditions for the poly(aryl benzyl ether) and poly-
(aryl benzyl thioether) dendrons,5a,11b we assume that a complete
conversion of the hydroxyl terminal groups into (diphenylphos-
phino)benzoxy moieties was achieved for all generations of the
polyamine dendron. Quantification of the phosphine loading was
achieved using the 31P NMR. Each resin was mixed with
commercial polystyrene-immobilized triphenylphosphine refer-
ence resin with a known loading. After the 31P gel-phase NMR
spectra were recorded, the yields were determined using the
mixing ratio, the integral ratio, and the known reference
loading.16 More accurate results were obtained after the resins
were allowed to swell in benzene-d6 for 24 h. Both strategies
usually showed an excellent agreement.
The Heck Reaction. To examine the dependence of the
performance of polymer-bound catalytic systems on the dendritic
architecture of the spacer, tethering the catalytic units to the
matrix, the Heck reaction was chosen. Heck olefin arylation,
one of the most widely used metal-catalyzed processes in
synthetic organic chemistry, was successfully accomplished in
solution with aryl iodides, bromides, and even chlorides, using
a variety of catalytic systems, particularly phosphine-palladium
complexes.17 Heterogeneous catalysis, however, was performed
almost entirely with iodides or electron-deficient bromides,
mainly using metal palladium adsorbed on an inorganic
support.18-20 Only a few phosphine-based heterogeneous sys-
tems have been reported, and the reactivities of these systems
were almost always limited to iodides.21 The fundamental study
by Hallberg and co-workers exposed the need for multiple
phosphine ligation to Pd as a tool for effective bromoarene
a Reagents and conditions: (i) 2,6-lutidine, TBAI, 5, DMF, 80 °C, 3
days; (ii) LiBH4, B(OMe)3, THF, 67 °C, 12 h; (iii) SO3·Py, NEt3, DMSO,
4 h, room temperature; (iv) NaBH(OAc)3, HC(OMe)3, 5, DMF, 60 °C, 48
h.
the dendron is assembled via the reductive amination of the
aldehydes with 5 in the presence of trimethylorthoformate, using
sodium triacetoxyborohydride as a reducing agent. Apparently,
the nucleophilicity of the aniline monomer is sufficient for
formation of the transient imine conjugate during the reductive
amination step. Repetitive ester reduction, oxidation, and
reductive amination sequence led to the formation of the higher
generations. We found that this synthetic scheme can be cleanly
and effectively accomplished, yielding in eight steps, starting
from Wang Bromo resin, 85% of the third generation dendron.
The synthesis was monitored, and the products characterized,
using gel-phase 13C NMR as well as acidolytic cleavage,
1
followed by H and 13C NMR and mass spectrometry. Similar
to the polythioether dendron, the cleavage always occurred
between the resin and the Wang linker, thus forming the
4-hydroxybenzyl-protected version of the dendron (Scheme 4b).
Again, similar to the polythioether case, the linker is clearly
1
visible in the H and 13C NMR spectra upon cleavage, but is
(15) NoVabiochem-Catalog and Peptide Synthesis Handbook; Novabiochem:
Laufelfingen, 1999; p S42.
removed under CI-MS conditions.
(16) Bar-Nir Ben-Aroya, B.; Portnoy, M. Tetrahedron 2002, 58, 5147.
(17) Beletskaya, I. P; Cheprakov, A. V. Chem. ReV. 2000, 100, 3009.
(18) (a) Kivaho, J.; Hanaoka, T.; Kubota, Y.; Sugi, Y. J. Mol. Catal. 1995,
101, 25. (b) Zhao, F.; Bhanage, B. M.; Shirai, M.; Arai, M. Chem.-Eur. J.
2000, 6, 843. (c) Biffis, A.; Zecca, M.; Basato, M. Eur. J. Inorg. Chem.
2001, 1131. (d) Davies, I. W.; Matty, L.; Hughes, D. L.; Reider, P. J. J.
Am. Chem. Soc. 2001, 123, 10139. (e) Zhao, F.; Shirai, M.; Ikushima, Y.;
Arai, M. J. Mol. Catal. A 2002, 180, 211.
Use of the amine linkage to the support required a slight
change in the procedure for cleavage of the dendrons. We found
that, as expected, only a low percentage of the dendrons were
cleaved by a TFA/CDCl3 (1:1 v/v) solution after 1 h of stirring.
However, the dendrons were quantitatively released from the
resin following 24 h cleavage.
(19) For rare metal palladium catalysis capable of unactivated bromoarene
olefination, see: (a) Ko¨hler, K.; Wagner, M.; Djakovitch, L. Catal. Today
2001, 66, 105 and references therein. (b) Mehnert, C. P.; Ying, T. Y. Chem.
Commun. 1997, 2215. (c) Ko¨hler, K.; Heidenreich, R. G.; Krauter, J. G.
E.; Pietsch, J. Chem.-Eur. J. 2002, 8, 622.
(20) For rare examples of supported systems capable of unactivated bromoarene
olefination, see: (a) Buchmeiser, M. R.; Wurst, K. J. Am. Chem. Soc. 1999,
121, 11101. (b) Schwarz, J.; Bo¨hm, V. P. W.; Gardiner, M. G.; Grosche,
M.; Herrman, W. A.; Hieringer, W.; Raudaschl-Sieber, G. Chem.-Eur. J.
2000, 6, 1773. (c) Dell’Anna, M. M.; Mastrorilli, P.; Muscio, F.; Nobile,
C. F.; Surranna, G. P. Eur. J. Inorg. Chem. 2002, 1094.
Although characterization of the products of the first genera-
tion was easily accomplished using gel-phase 13C NMR and/or
1
acidolytic cleavage, followed by H and 13C NMR and mass
spectrometry, it was much more difficult to characterize the
products of the second and the third generations. Partial
protonation of the amines in the cleavage solution caused severe
difficulties in assignment of the peaks in the 1H and 13C NMR.
In addition, line broadening by neighboring nitrogens interfered
in the assignment of the gel-phase NMR spectra.
(21) (a) Wang, P.-W.; Fox, M. A. J. Org. Chem. 1994, 59, 5358. (b) Villemin,
D.; Jaffre´s, P. A.; Nechab, B.; Courivaud, F. Tetrahedron Lett. 1997, 38,
6581. (c) Riegel, N.; Darcel, C.; Ste´phan, O.; Juge´, S. J. Organomet. Chem.
1998, 567, 219.
9
J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007 5863

A R T I C L E S
Scheme 7
Dahan and Portnoy
Table 1. Heck Reaction with Methyl Acrylate at 80 °C for 72 ha
entry
catalyst
conversion (%)
yield of 10 (%)
selectivity (10:11)
1
2
3
4
5
6
7
8
9
6
20
11
66
100
15
19
67
18
25
91
84
18
11
66
100
10
13
61
13
18
89
82
9.0
22
>130
>200
2
7G1
7G2
7G3
8G1
8G2
8G3
9G1
9G2
9G3
12
olefination with a supported catalyst.22 Unfortunately, such
coordination modes result in severe cross-linking of the polymer,
which frequently has a negative effect on the catalytic activity.
Another work, performed in our group, demonstrated the clear
advantage of the bidentate over monodentate phosphine ligands
in the case of the heterogeneously catalyzed Heck reaction.23
With this notion in mind, we decided to investigate catalytic
systems that are based on phosphine-palladium complexes
derived from dendronized polymeric supports.
2.2
10.2
2.6
2.6
44.5
41
10
11
a Reagents and conditions: 0.5 mmol of bromobenzene, 0.6 mmol of
methyl acrylate, 0.65 mmol of triethylamine, and the catalyst (0.0125 mmol
of Pd) in 1 mL of N-methylpyrrolidone (NMP), 80 °C, 72 h.
on the matrix, this is counterbalanced by the increase in the
mass of the spacer and some decrease in the overall synthetic
yield.
The reaction of bromobenzene with methyl acrylate, chosen
as a model for the Heck reaction with catalysts 6-9, was
performed under two different temperature regimes: at 80 °C
for 72 h and at 120 °C for 14 h (eq 1). While the main product
is trans-methyl cinnamate (10), in some experiments the
bromobenzene homocoupling byproduct, biphenyl (11), was
obtained.
The dendronized resins were decorated with phosphine groups
via the attachment of 4-(diphenylphosphino)benzoic acid to the
periphery of the dendrons, as explained for the polyamine
dendrons characterization (Scheme 7). The resins were char-
acterized using gel-phase 31P NMR (signal at -4.2 ppm) and
1
13C NMR as well as TFA-induced cleavage followed by H
NMR of the cleaved solution. Quantitative determination of the
1
resin loading based on both 31P and H NMR usually showed
excellent agreement. Complete conversion of the hydroxyl
terminal groups into (diphenylphosphino)benzoxy moieties was
achieved for all resins.
Study of the properties of the Wang(PPh2) and GnX(PPh2)
resins revealed that THF is the best solvent for their swelling,
and, accordingly, it was chosen for their complexation with
metal precursors. Thus, incubation of the phosphine-terminated
resins with Pd(dba)2 in deoxygenated THF at ambient temper-
ature for 4 h yielded reddish brown resins 6-9 (Scheme 8).
Gel-phase 31P NMR demonstrated quantitative complexation of
the phosphines with Pd (signals at 25-30 ppm). 1H NMR
spectra of the TFA-cleaved solutions of resins 6-9 revealed
that, for every two phosphine groups, one molecule of diben-
zylidenacetone is present. Thus, the most probable structure of
the Pd complexes on the supports is (phosphine)2Pd(dba), and
the loading of the active catalyst on support was calculated
accordingly.24 As to these calculations, the loading of Pd for
resins 6-9 ranges between 0.3 and 0.5 mmol/g resin. Although
upon increase of the dendron generation, there is an increase in
the amount of the complexation sites per single attachment point
The results of the model reaction, of the three types of
dendrons at 80 °C for 72 h, summarized in Table 1 and Chart
1, demonstrate a remarkable improvement in the catalytic
performance upon dendronization of the support for all types
of dendrons, and the importance of the character of the dendron
backbone on the performance.
The difference in the reactivity within the 7 series of catalysts
is very dramatic under these conditions (Table 1, entries 1-4).
While there is a notable decrease in the activity of 7G1, as
compared to that of 6, the activity increases significantly for
7G2, improves further for 7G3, surpassing by far that of 6, and
rendering for 7G3 quantitative conversion and yield. It is
noteworthy that, similar to 6 and 7G1, most of the known
catalytic systems for the Heck reaction of bomoarenes are not
Scheme 8
9
5864 J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007

Pd Catalysis on Dendronized Solid Support
A R T I C L E S
Chart 1. Catalytic Results of the Heck Reaction at 80 °C:
Table 2. Heck Reaction with Methyl Acrylate at 120 °C for 14 ha
Comparison of the Three Types of Dendrons
entry
catalyst
conversion (%)
yield of 10 (%)
selectivity (10:11)
1
2
3
4
5
6
7
8
9
6
95
100
100
100
56
45
93
100
100
100
75
81
86
97
51
37
89
90
93
93
3.8
4.3
6.1
32
10.2
4.6
22.2
9
13.3
13.3
7G1
7G2
7G3
8G1
8G2
8G3
9G1
9G2
9G3
10
a Reagents and conditions: 0.5 mmol of bromobenzene, 0.6 mmol of
methyl acrylate, 0.65 mmol of triethylamine, and the catalyst (0.0125 mmol
of Pd) in 1 mL of NMP, 120 °C, 14 h.
active enough at 80 °C and require higher temperatures (110-
130 °C). The selectivity of the catalysis also improves signifi-
cantly upon the increase of the dendron generation, and no
biphenyl is observed for 7G2 and 7G3. We also synthesized
the non-dendritic homogeneous analogue 12 of complexes 7,
prepared in situ from Pd(dba)2 and the methyl ester of the
4-(diphenylphosphino)benzoic acid. Remarkably, the conversion
and selectivity achieved with 7G3 (Table 1, entry 4) are
significantly higher than those obtained with 12 (Table 1, entry
11). One may argue that 12 incorporates monodentate phos-
phines, while those forming 7 are chelating in nature. However,
the minimal chelate size on the dendronized resins is 20 atoms,
the tether between every two phosphine groups is fairly flexible,
and, thus, two resin-bound phosphines coordinated to Pd in 7
are very similar in nature to the monodentate ligands. Therefore,
chelate effects are unlikely to be responsible for the superiority
of 7G3 over 12.
Chart 2. Catalytic Results of the Heck Reaction at 120 °C:
Comparison of the Three Types of Dendrons
Table 3. Heck Reaction with Acrylonitrilea
entry
catalyst
conversion (%)
yield of cinnamnitrile 13 (%)
selectivity (13:11)
1
2
3
4
6
53
49
73
76
49
46
68
72
12
15
14
18
7G1
7G2
7G3
a Reagents and conditions: 0.5 mmol of bromobenzene, 0.6 mmol of
acrylonitrile, 0.65 mmol of triethylamine, and the catalyst 7 (0.0125 mmol
of Pd) in 1 mL of NMP, 120 °C, 14 h.
The performance of the polythioether-derived catalysts is
inferior to those based on the other two dendron types, also at
120 °C (Chart 2). These results show that the backbone of the
polythioether dendron (8) is, to some extent, a catalytic poison,
as we assumed at the beginning of our work. The differences
between the polyether and polyamine dendron-based catalysts
7 and 9 are very significant at low temperatures, but at high
temperatures the differences were hardly observable, and it
seems that, in some cases, the polyamine dendron is even
slightly more active (Chart 2). It can be concluded that, at low
temperatures, the competition between the heteroatom (nitrogen
or sulfur), located in the dendron backbone, and the reacting
olefin dominates the reaction outcome, while less so at higher
temperatures.
Similar results were obtained for the 8 and 9 series. The
catalysts derived from the third generation are superior to the
non-dendronized catalysts in all parameters.
Although there are no significant differences between the
catalysts derived from the three types of the first generation
dendron (7G1, 8G1, 9G1), a dramatic difference is observed
for those based on the second generation (7G2, 8G2, 9G2),
where 7G2 is a much more active and selective catalyst than
the other two (Chart 1). This pattern is preserved for the third
generation-derived catalysts (7G3, 8G3, 9G3).
The results of the model reaction with the three types of
dendritic catalysts at 120 °C are summarized in Table 2 and
Chart 2. In the case of the Heck reaction at 120 °C, the pattern
for the 7 series follows the trends observed in the former case.
Although, under these conditions, the conversions obtained with
the dendritic catalysts are only marginally better than that of
the catalyst derived from the Wang resin (6), the selectivity for
the formation of the Heck product consistently improves from
6 to 7G3 (an 8-fold increase is achieved, Table 2, entries 1 and
4). This increase in selectivity is reflected in the substantially
higher yield reached using the third generation-derived 7G3,
as compared to its nondendronized analogue.
This explanation is further supported by another experiment
we performed. In this experiment, we used 7, as catalysts, in
the Heck arylation of acrylonitrile with bromobenzene (eq 2,
Table 3).
At 80 °C, only traces of the product are observed, and it is
hard to tell whether there is any dendritic effect. However, at
120 °C, the conversions and yields are much higher, and a
prominent dendritic effect is visible. The substrate and product
(22) Andersson, C.-M.; Karabelas, K.; Hallberg, A.; Andersson, C. J. Org. Chem.
1985, 50, 3891.
(23) Mansour, A.; Portnoy, M. Tetrahedron Lett. 2003, 44, 2195.
(24) For similar (phosphine)2Pd(dba) complexes on dendronized silica, see:
Zweni, P. P.; Alper, H. AdV. Synth. Catal. 2004, 346, 849.
9
J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007 5865

A R T I C L E S
Dahan and Portnoy
Table 4. Heck Reaction with Styrenea
Table 5. The Suzuki Reaction with 4-Methoxyphenylboronic Acida
entry
catalyst
conversion (%)
yield of trans stilbene 14 (%)
selectivity (14:11)
entry
catalyst
conversion/yield (%)
1
2
3
4
5
6
7
8
6b
99
73
67
56
81
89
66
67
68
91
2.0
3.3
4.3
8.1
1.9
2.0
2.1
10
1
2
3
4
5
6
7
8
9
6
41
44
78
84
50
56
82
57
44
78
7G1b
7G2b
7G3b
6c
7G1
7G2
7G3
8G1
8G2
8G3
9G1
9G2
9G3
100
100
100
100
100
100
7G1c
7G2c
7G3c
a Reagents and conditions: 0.5 mmol of bromobenzene, 0.6 mmol of
styrene, 0.65 mmol of triethylamine, and the catalyst 7 (0.0125 mmol of
Pd) in 1 mL of NMP. b 80 °C, 72 h. c 120 °C, 14 h.
10
a Reagents and conditions: 0.5 mmol of bromobenzene, 0.6 mmol of
4-methoxyphenylboronic acid, 0.65 mmol of triethylamine, and the catalyst
(0.0125 mmol of Pd) in 1 mL of NMP, 80 °C, 14 h.
can act as ligands, by themselves, coordinating to Pd through
the nitrile group. Thus, the competition between the nitrile and
the olefin coordination leads to a very low reactivity at 80 °C.
At 120 °C, the kinetics is much faster, and the aforementioned
competition is of lesser importance.
Chart 3. Catalytic Results of the Suzuki Reaction at 80 °C:
Comparison of the Three Types of Dendrons
When styrene is examined as a substrate (eq 3), the trends in
activity and selectivity of the catalysts 7 are very similar to
those observed with methyl acrylate (Table 4), although the
magnitude of the effect is somewhat smaller.
reactions.27 While both reactions start with the oxidative addition
step, only in the case of the Heck reaction does the second step
involve coordination of a weak ligand (olefin) to the metal prior
to the formation of the C-C bond. In the case of the Suzuki
reaction, the second step is the transmetallation, and there is
no subsequent step in which a weak ligand (such as an olefin)
has to compete for the coordination site on the metal. It is
possible that, for this reason, the interference from the heteroa-
toms located in the dendron backbone is less significant in the
Suzuki reaction.
The Nature of the Active Catalyst. Precipitation of metallic
palladium occurs during the reaction, and, thus, the possibility
of the catalysis being performed by Pd metal nanoparticles,
stabilized inside the dendritic matrix, had to be considered.28 It
is well known that metal Pd adsorbed on an insoluble support
or dendrimer-stabilized Pd nanoparticles catalyze the Heck
reaction of aryl iodides or electron-withdrawing group-activated
bromides.18,29 However, these catalytic entities are usually
incapable of functionalizing nonactivated aryl bromides (such
as bromobenzene in this study).19 Indeed, whereas at the end
of the catalytic reaction the resin and the precipitated Pd are
easily separated by simple filtration, this solid exhibited a
substantially diminished catalytic activity in the recycling
experiment. Additional data, ruling out the possibility of the
metal palladium catalysis, were obtained in the experiment, in
which the Heck reaction with the electron-rich butyl vinyl ether
was tested with catalysts 6 and 7 (Scheme 9, Table 6).
For this reactant, the regioselectivity and steroselectivity are
important parameters, as three products, cis- and trans-1,2-
disubstituted 17 and 18, as well as 1,1-disubstituted 16
(undergoing hydrolysis to acetophenone 19 under the reaction/
workup conditions), are formed. Moreover, the regio- and
The Suzuki Reaction. The Suzuki reaction, the coupling of
aryl halides with aryl boronic acids to form symmetrically or
non-symmetrically substituted biaryls, is another popular method
for C-C bond formation.25 Like the Heck reaction, the Suzuki
reaction is generally carried out using palladium phosphine
complexes. Suzuki reaction, using Pd supported on active
charcoal, silica, or micro- and mesoporous aluminosilicates as
heterogeneous catalysts, has been reported.26
The Suzuki reaction was chosen as another process to be
examined with supported dendronized palladium catalysts 7-9.
For this purpose, we used the reaction of bromobenzene and
4-methoxyphenylboronic acid as a model (eq 4). The conversion
and yield, in the case of this reaction, are identical, because
4-methoxybiphenyl (15) is the only product.
The results of the reaction with the three types of dendrons
at 80 °C, summarized in Table 5 and Chart 3, demonstrate a
significant improvement in the catalytic performance upon
dendronization of the support for all types of dendrons. They
also show that the character of the dendron is less significant,
as for the third generation the three types of dendrons showed
very similar reactivity (Table 5, entries 4, 7, and 10).
A possible explanation of this observation arises from
examination of the mechanisms of the Heck and Suzuki
(25) (a) Kotha, S.; Lahiri, K.; Kashinath, D. Tetrahedron 2002, 58, 9633. (b)
Bai, L.; Wang, J. X. Curr. Org. Chem. 2005, 9, 535.
(26) (a) Bai, L.; Wang, J. X. Curr. Org. Chem. 2005, 9, 535. (b) Uozumi, Y.
Top. Curr. Chem. 2004, 242, 77. (c) Brase, S.; Kirchhoff, J. H.; Kobberling,
J. Tetrahedron 2003, 59, 885. (d) Paetzold, E.; Oehme, G.; Fuhrmann, H.;
Richter, M.; Eckelt, R.; Pohl, M.-M.; Kosslick, H. Microporous Mesoporous
Mater. 2001, 44/45, 517. (e) Kosslick, H.; Mo¨nnich, I.; Paetzold, E.;
Fuhrmann, H.; Fricke, R.; Mu¨ller, D.; Oehme, G. Microporous Mesoporous
Mater. 2001, 44/45, 537.
(27) Hegedus, L. S. In Organometalics in Organic Synthesis; Schlosser, M.,
Ed.; Wiley: Chichister, 1994; p 383.
(28) Niu, Y. H.; Crooks, R. M. C. R. Chem. 2003, 6, 1049.
(29) Yeung, L. K.; Crooks, R. M. Nano Lett. 2001, 1, 14.
9
5866 J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007

Pd Catalysis on Dendronized Solid Support
A R T I C L E S
Scheme 9
a similar phenomenon for partially complexed resin-bound
phosphines in the carbonylation of bromobenzenes.31 Thus, there
is a possibility that, in our experiments with dendronized
catalysts of higher generations, more active catalytic species
are formed upon partial precipitation of palladium (because
precipitation frees the phosphine sites), and these species are
responsible for the decrease in the â-arylation/R-arylation ratio.
Even more significant is the ratio between products 18 (trans)
and 17 (cis) (Table 7), which is known to be sensitive to the
nature of the catalyst. According to the literature,22 phosphine-
Pd complexes as catalysts yield ca. a 2.5:1 ratio, while the
“naked” Pd catalysts form a 1:1 product mixture. The ratio
produced in our experiments (2.4-3.0 to 1) is much more
characteristic of the phosphine-palladium catalysts.
Possible Sources of the Dendritic Effects. Positive dendritic
effects, in general, and those in supported catalysis, in particular,
seem to be complex phenomena, being an expression of various
characteristic properties of the dendritic systems.7 However, in
the case at hand, we can point to two basic factors likely to
play a role in inducing the positive effects. The first and,
probably, dominant property indirectly responsible for the effects
is the increased proximity of the ligating sites (higher local
loading of the ligands). This property is translated into reduced
cross-linking of polystyrene. The reduced cross-linking in our
systems was observed with both the naked eye and NMR (a
narrowing of signals in the gel-phase 31P NMR) and may lead
to improved swelling, a more solvent-like environment for
catalytic sites, and improved reagent transport and product
removal.
Table 6. Heck Reaction with Butyl Vinyl Ethera
combined yield of
selectivity
17 18]:11)
entry
catalyst
conversion (%)
16
+
17
+
18 (%)
([16
+
+
1
2
3
4
5
6
7
8
6b
55
49
54
46
41
46
72
58
58
75
83
5.1
5.1
5.8
7G1b
7G2b
7G3b
6c
77
14
100
100
100
100
1.4
1.4
3.0
4.9
7G1c
7G2c
7G3c
a Reagents and conditions: 0.5 mmol of bromobenzene, 0.6 mmol of
butyl vinyl ether, 0.65 mmol of triethylamine, and the catalyst (0.0125 mmol
of Pd) in 1 mL of NMP. b 80 °C, 72 h. c 120 °C, 14 h.
Table 7. Regio- and Stereoselectivity in the Heck Reaction of
Butyl Vinyl Ether
yield of
â
+
-arylation
yield of
R
-arylation
regioselectivity
([17 18]/[19])
stereoselectivity
catalyst
(17
18, %)
(19, %)
+
(18/17)
6a
37
30
29
44
33
31
40
42
9
10
17
28
25
27
35
41
4.1
2.6
2.6
2.8
3.0
2.4
2.6
3.0
3.0
7G1a
7G2a
7G3a
6b
3.0
1.7
1.6
1.3
1.2
1.1
1.0
Another expression of the increased proximity of the ligating
sites contributing to the positive effect is, most likely, the
preorganization of the site to form a more effective catalytic
unit through the release of free phosphine arms upon partial
precipitation of Pd, as discussed above. The effect of partial
complexation exists even in systems with random distribution
of the ligating sites22,31 and must be strengthened when the free
ligands are preorganized in the complex proximity.
7G1b
7G2b
7G3b
1
a 80 °C, 72 h. b 120 °C, 14 h. c Determined by H NMR.
stereoselectivity in this reaction can provide important mecha-
nistic input, as described below. The trends observed for the
conversion to Heck products (16-18) and selectivity toward
these products (versus 11) are very similar to those obtained
for methyl acrylate and styrene. At both temperatures, 7G3 is
clearly superior to 6.
We previously demonstrated, for a different system, that
increased proximity of the ligating sites can lead to differences
in the population of the metal species on the resin.5a However,
this is not the case for the current system because, as
aforementioned, even for non-dendronized support, we found
that two ligands coordinated to a single Pd(dba) fragment is
the dominant species.
According to the regio- and stereoselectivity data (sum-
marized in Table 7), the proportion of the â-arylated products
17 + 18 decreases as the generation of the dendritic template
increases. The decrease in the ratio â-arylation/R-arylation for
butyl vinyl ether can be attributed to the higher portion of the
olefin insertion occurring through cationic intermediates (en-
forcing R-arylation) versus neutral intermediates (yielding a
mixture of R- and â-arylated enol ethers).30 This change in the
distribution between the two alternative catalytic pathways may
result from the more polar environment of the catalytic units in
higher generation catalysts. The reason for this change may also
be associated with the increased local loading of free phosphines
in higher generation dendrons. In recent work performed in our
group, it was found that partial, rather than full, complexation
of the tridentate phosphine ligand with a Pd(dba) fragment
(complexation, leaving a free phosphine arm) led to a more
active and selective Heck catalyst, which also induced a higher
proportion of R- to â-arylation of butyl vinyl ether. We observed
The second property that is substantially different for the
dendronized, versus the non-dendronized support, is the length
of the spacer between the anchoring site on the polymer core
and the catalytic unit. Longer spacers may lead to the increased
distance between the core and the catalytic moiety, placing the
latter into the more solvent-like environment. To examine
whether the longer spacer is the key to the improved perfor-
mance of the higher generation-derived catalysts in our systems,
we synthesized resin 20 starting from Wang trichloroacetimidate
resin (Scheme 10). We assumed that the length of the spacer of
the ligands in resin 20 is similar to that of 7G3.
Resin 20 was complexed with Pd(dba)2 under the conditions
previously mentioned, yielding supported complex 21. Gel-phase
31P NMR demonstrated quantitative complexation of the phos-
(30) Cabri, W.; Candian, I. Acc. Chem. Res. 1995, 28, 2.
(31) Mansour, A.; Portnoy, M. J. Mol. Catal. A: Chem. 2006, 250, 40.
9
J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007 5867

A R T I C L E S
Scheme 10a
Dahan and Portnoy
Pd precipitated during the processes is, most probably, not
involved in the catalysis.
Analysis of the possible sources of the dendritic effect pointed
to the increased proximity of the ligands, translated into reduced
cross-linking of the polymer and preorganization of the ligating
site, as the main property leading to the increased activity and
selectivity of the higher generation catalysts. It was demonstrated
that the length of the spacer, on its own, is not the source of
the effect.
Catalysis dependence on the nature of the dendritic backbone
was exposed. This backbone influence is stronger for the Heck
reaction at lower temperatures, with the polyether backbone
being the optimal one. The backbone influence is smaller for
the Heck reaction at higher temperatures and for the Suzuki
coupling. Competition of the backbone coordinating elements
with the substrate is suggested as the reason for the observed
behavior.
a Reagents and conditions: (i) BF3·OEt, DCM, cyclohexane, room
temperature, 10 min; (ii) Cs2CO3, KI, DMF, 70 °C, 24 h.
Chart 4. The Influence of the Length of the Catalyst Spacer on
the Heck and Suzuki Reactions
Experimental Section
General Information. All reactions were conducted under an
atmosphere of nitrogen in oven-dried glassware with magnetic stirring.
1H NMR and 13C NMR spectra were recorded on Bruker AVANCE-
200 and AVANCE-400 spectrometers. MS and MALDI-TOF MS were
recorded on Micromass VG-Autospec M250 and Bruker Reflex-3 mass
spectrometers or a Voyager DE-STR, respectively.
phine ligands with Pd. The catalytic results of 21 were compared
to those of 6 and 7G3 in the two test reactions as shown in
Chart 4.
1
Yields were determined using the H NMR of TFA:CDCl3 (1:1)
cleavage solutions with C6H6 (7.36 ppm) as an internal standard.
Alcohols were partially converted to TFA esters under these conditions.
The syntheses of poly(aryl benzyl ether) and poly(aryl benzyl
thioether), functionalization of Wang and GnO(OH) resins with
diphenylphosphinobenzoic acid, and a general procedure for the Heck
reaction with bromobenzene were previously communicated.5a,b,11
Synthesis of G1N(CO2Me) (Attachment of Dimethyl 5-Ami-
noisophthalate to Wang Resin). Dimethyl 5-aminoisophthalate (0.45
g, 2.2 mmol, 3 equiv), 2,6-lutidine (0.25 mL, 2.2 mmol, 3 equiv), and
tetrabutylammonium iodide (0.8 g, 2.2 mmol, 3 equiv) were added to
a suspension of Bromo Wang resin (1.0 g, 0.72 mmol/g, 0.72 mmol, 1
equiv) in DMF (10 mL). The suspension was mixed at 80 °C for 3
days. The resin was washed with DMF (3 × 20 mL), DMF/water (1:1,
3 × 20 mL), THF/water (1:1, 3 × 20 mL), THF (3 × 20 mL), and
dichloromethane (3 × 20 mL), and then dried under vacuum. Yield
>99%, purity >99%, loading 0.66 mmol/g. Partial gel-phase 13C NMR
(100.8 MHz, C6D6): δ 165.9 C5, 153.2 C11, 148.4 C1, 131.0 C3,
117.2 C2 + C4, 114.6 C10, 69.1 CMerrifield, 51.2 C6, 40.1 C7. Following
TFA-induced cleavage: 1H NMR (200 MHz, CDCl3/TFA 1:1) δ 8.86
(s, 1H, H2), 8.48 (s, 2H, H1), 7.23 (d, J ) 7.0 Hz, 2H, H5), 6.93 (d,
J ) 7.0 Hz, 2H, H6), 4.73 (s, 2H, H4), 4.09 (s, 6H, H3). Partial 13C
NMR (100.8 MHz, CDCl3/TFA 1:1): δ 166.9 C5, 132.8 C9, 132.6
C4, 130.2 C3, 129.4 C2, 115.9 C10, 54.1 C6. MS (EI) found (m/z),
209.1; calcd for C10H11NO4, 209.2.
The catalytic results obtained with complex 21 and those
obtained with complex 6 are very similar, while the reactivity
of catalyst 7G3 is much higher. From these results, we can
conclude that the length of the spacer between the core and the
catalytic unit is, by itself, not the cause of the increased
reactivity, although it is possible that, when combined with a
dendritic architecture, the longer spacer indeed increases the
average distance of the catalytic units from the core.
A very recent series of experiments, where GnO resins were
functionalized with bidentate chelating phosphine ligands,
demonstrated a negative effect in the Heck reaction.32 These
data strongly support the preorganization of the ligating sites
and reduced cross-linking, rather than increased average distance
of the units from the core, as the main reasons for the positive
effects reported here.
Conclusions
This study established effective and clean routes to the
construction of poly(aryl benzyl ether), poly(aryl benzyl thio-
ether), and poly(aryl benzyl amine) dendrons on polystyrene
beads from readily available starting materials. These synthetic
routes led to dendronized supports with up to third generation
dendrons grafted on the polymer core. These supports, once
functionalized with phosphine-palladium complexes, exhibited
remarkable dendritic effects in the Heck and Suzuki reactions
of bromobenzene. The generation dependence is most notable
for the Heck reaction at 80 °C, unusually mild conditions for
this reaction, and results in a fivefold increase in yield for the
reaction with methyl acrylate, when the best dendritic catalyst
(third generation 7G3) is compared to the non-dendritic
analogue. Experimental and literature data point to the phos-
phine-palladium complexes as the active catalysts, while the
Typical Procedure for the Reduction Step. Synthesis of G1N-
(OH). LiBH4 (6.6 mL, 13 mmol, 20 equiv, 2 M in THF) and B(OMe)3
(0.074 mL, 0.66 mmol, 1 equiv) were added to a suspension of the
resin G1N(CO2Me) (1.0 g, 0.66 mmol/g, 0.66 mmol, 1.3 mmol groups
of ester, 1 equiv) in THF (10 mL). The mixture was refluxed overnight.
(32) Mansour, A. Supported Catalytic Systems Based on Aminoalcohol-Derived
Phosphine Ligands. Ph.D. Thesis, Tel Aviv University, Tel Aviv, 2005.
9
5868 J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007

Pd Catalysis on Dendronized Solid Support
A R T I C L E S
The resin was washed with water (3 × 20 mL), a solution of ammonium
chloride/THF (1:1, 3 × 20 mL), a solution of 10% HCl/THF (1:1, 3 ×
20 mL), water (3 × 20 mL), THF (3 × 20 mL), dichloromethane (3 ×
20 mL), and then dried under vacuum. Yield >95%, purity >95%,
loading 0.65 mmol/g. Partial gel-phase 13C NMR after acylation of the
OH and NH groups with acetyl chloride (100.8 MHz, C6D6): δ 175.9
C13, 169.1 C11, 143.2 C1, 137.8 C3, 130.0 C9, 114.3 C7, 69.4
(s, 12H, H6). Partial 13C NMR (100.8 MHz, CDCl3/TFA 1:1): δ 166.0
C10, 132.7 C9, 132.3 C4, 132.1 C2, 129.8 C8, 128.5 C7, 116.2 C14,
53.8 C11. HRMS (EI): found (m/z), 535.1943; calcd for C28H29N3O8,
535.1950.
CMerrifield, 64.5 C5, 51.6 C6, 22.1 C12, 19.8 C10. Following TFA-
induced cleavage: 1H NMR (200 MHz, CDCl3/TFA 1:1) δ 7.60 (s,
1H, H2), 7.44 (s, 2H, H1), 7.16 (d, J ) 8.2 Hz, 2H, H5), 6.90 (d, J )
8.2 Hz, 2H, H6), 5.39 (s, 4H, H3), 4.63 (s, 2H, H4). Partial 13C NMR
(100.8 MHz, CDCl3/TFA 1:1): δ 138.0 C3, 132.9 C8, 130.4 C4, 124.1
C2, 117.0 C9, 68.4 C5. HRMS (EI): found (m/z), 345.0439; calcd for
C12H9F6NO4, 345.0436.
General Procedure for Complexation of Resins Wang(PPh2),
GnX(PPh2), and 20 with Pd(dba)2. Syntheses of 6-9 and 21. In a
glove box, a solution of Pd(dba)2 in THF (0.7 equiv/phosphine group,
0.5 mmol/10 mL of THF) was added to the phosphine-bearing resin
swollen in THF (10 mL/g resin). The suspension was gently stirred
for 4 h. The product resin was filtered, washed with THF and ether,
and dried under vacuum.
6. Loading of palladium: 0.38 mmol/g. 31P NMR (162 MHz,
C6D6): 24.9.
7G1. Loading of palladium: 0.36 mmol/g. 31P NMR (162 MHz,
C6D6): 25.9.
7G2. Loading of palladium: 0.47 mmol/g. 31P NMR (162 MHz,
C6D6): 25.9.
7G3. Loading of palladium: 0.31 mmol/g. 31P NMR (162 MHz,
C6D6): 27.0.
8G1. Loading of palladium: 0.37 mmol/g. 31P NMR (162 MHz,
C6D6): 25.2.
8G2. Loading of palladium: 0.51 mmol/g. 31P NMR (162 MHz,
C6D6): 25.9.
Typical Procedure for the Oxidation Step. Synthesis of G1N-
(CHO). Sulfur trioxide pyridine complex (1.8 g, 11 mmol, 17 equiv)
was added to a mixture of DMSO (7.5 mL) and triethylamine (7.5
mL). The mixture was mixed for 15 min at room temperature and then
added to a suspension of G1N(OH) (1.0 g, 0.65 mmol/g, 0.65 mmol,
1 equiv) in DMSO (10 mL). The suspension was mixed at room
temperature for 5 h. The resin was washed with DMSO (3 × 20 mL),
DMSO/water (3 × 20 mL), THF/water (3 × 20 mL), THF (3 × 20
mL), and dichloromethane (3 × 20 mL) and then dried under vacuum.
Yield >99%, purity >95%, loading 0.65 mmol/g. Partial gel-phase 13
C
NMR (100.8 MHz, C6D6): δ 190.7 C5, 148.8 C1, 145.3 C7, 137.9
C3, 121.2 C4, 116.5 C2, 114.9 C9, 69.7 CMerrifield, 46.9 C6. Following
TFA-induced cleavage: 1H NMR (200 MHz, CDCl3/TFA 1:1) δ 10.13
(s, 2H, H3), 8.72 (s, 1H, H2), 8.46 (s, 2H, H1), 7.23 (d, J ) 7.0 Hz,
2H, H5), 6.93 (d, J ) 7.0 Hz, 2H, H6), 4.73 (s, 2H, H4). Partial 13C
NMR (100 MHz, CDCl3/TFA 1:1): δ 193.0 C5, 138.0 C3, 134.1 C4,
131.8 C8, 130.3 C2, 116.0 C9. HRMS (EI): found (m/z), 149.0474;
calcd for C8H7NO2, 149.0477.
8G3. Loading of palladium: 0.52 mmol/g. 31P NMR (162 MHz,
C6D6): 26.0.
9G1. Loading of palladium: 0.40 mmol/g. 31P NMR (162 MHz,
C6D6): 26.1.
9G2. Loading of palladium: 0.50 mmol/g. 31P NMR (162 MHz,
C6D6): 27.0.
9G3. Loading of palladium: 0.44 mmol/g. 31P NMR (162 MHz,
C6D6): 30.0.
21. Loading of palladium: 0.23 mmol/g. 31P NMR (162 MHz,
C6D6): 25.9.
General Procedure for the Suzuki Catalysis. Bromobenzene (52
µL, 0.5 mmol, 1 equiv), 4-methoxyphenylboronic acid (91 mg, 0.6
mmol, 1.2 equiv), triethylamine (90 µL, 0.65 mmol, 1.3 equiv), and
one of the resins 6/7/8/9/21 (0.0125 mmol of Pd, 0.025 equiv) in NMP
(1 mL) in a pressure tube were sealed under nitrogen and heated to 80
°C for 14 h with stirring. After being cooled, the resin was filtered off
and washed with acetonitrile. The combined organic solution was
prepared for HPLC and NMR analysis.
Typical Procedure for the Reductive Amination Step. Synthesis
of G2N(CO2Me). Dimethyl 5-aminoisophtalate (1.3 g, 6.5 mmol, 10
equiv) and a catalytic amount of acetic acid were mixed in DMF until
full dissolution. The solution was added to G1N(CHO) (1.0 g, 0.65
mmol/g, 0.65 mmol, 1.3 mmol groups of aldehyde, 1 equiv), in DMF
(10 mL). Sodium triacetoxyborohydride (4.1 g, 19 mmol, 30 equiv)
and trimethyl orthoformate (1 mL) were added to the suspension of
the resin. The suspension was mixed at 60 °C for 48 h. The resin was
washed with DMF (3 × 20 mL), DMF/water (1:1, 3 × 20 mL), THF/
water (1:1, 3 × 20 mL), THF (3 × 20 mL), and dichloromethane (3 ×
20 mL) and dried under vacuum. Yield >99%, purity >95%, loading
0.52 mmol/g. Partial gel-phase 13C NMR (100.8 MHz, dioxane-d8): δ
169.4 C10, 131.3 C8, 120.5 C7 + C9, 55.3 C11, 44.2 C5. Following
TFA-induced cleavage: 1H NMR (200 MHz, CDCl3/TFA 1:1) δ 8.86
(s, 2H, H5), 8.47 (s, 4H, H4), 8.15-8.75 (br, 3H, H1 + H2), 6.70-
7.25 (broad, 4H, H8 + H9), 4.90-5.00 (broad, 6H, H3 + H7), 4.09
Acknowledgment. This research was supported by the Israel
Science Foundation. The fellowship of A.D. from the Ministry
of Science, Israel, is gratefully acknowledged.
Supporting Information Available: General experimental
conditions, procedures for the synthesis of 20, G2N, G3N, GnN-
(PPh2), and GnS(PPh2), and their characterization data. This
material is available free of charge via the Internet at
http://pubs.acs.org.
JA065265D
9
J. AM. CHEM. SOC. VOL. 129, NO. 18, 2007 5869