Synthesis of organometallic poly(dendrimer)s by macromonomer polymerization: effect of dendrimer size and structural rigidity on the polymerization efficiency

Article abstract of DOI:10.1002/chem.200802306

Two series of first to third generation (G1-G3) oligoether dendrimers, one hearing a shorter spacer chain (C-O) and the other having a longer spacer branch (C-C-C-O) were prepared. Both series of compounds, containing two reactive C≡CH moieties on the dendrimer surface, were used as macromonomers and copolymerized with trans-[Pt(PEt₃)₂Cl₂] to form organometallic poly(dendrimer)s by an outer-sphere-outer-sphere connection strategy. It was found that concentration of monomer used in the polymerization, the dendrimer generation, and, most strikingly, the length of the spacer were key factors that determined the polymerization efficiency. Hence, the structurally more rigid and compact C-O linked dendrimers formed poly(dendrimer)s with a higher degree of polymerization than the structurally less rigid and more bulky C-C-C-O dendrimers. This result was due to the higher tendency to form cyclic oligomers in the latter series of compounds. In addition, the differences in the polymerization efficiency among the three generations of dendrimers could be explained by the gradual decrease of reactive functional group density on the dendrimer surface.

Full text of DOI:10.1002/chem.200802306

DOI: 10.1002/chem.200802306
Synthesis of Organometallic Poly(dendrimer)s by Macromonomer
Polymerization: Effect of Dendrimer Size and Structural Rigidity on the
Polymerization Efficiency
Siu-Yin Cheung,[a] Hak-Fun Chow,*[a, b] To Ngai,[a] and Xiaoling Wei[a]
2278
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2278 – 2288

FULL PAPER
Abstract: Two series of first to third
generation (G1–G3) oligoether den-
drimers, one bearing a shorter spacer
concentration of monomer used in the
polymerization, the dendrimer genera-
tion, and, most strikingly, the length of
the spacer were key factors that deter-
mined the polymerization efficiency.
Hence, the structurally more rigid and
structurally less rigid and more bulky
ꢁ ꢁ ꢁ
C C C O dendrimers. This result was
due to the higher tendency to form
cyclic oligomers in the latter series of
compounds. In addition, the differences
in the polymerization efficiency among
the three generations of dendrimers
could be explained by the gradual de-
crease of reactive functional group
density on the dendrimer surface.
ꢁ
chain (C O) and the other having a
ꢁ ꢁ ꢁ
longer spacer branch (C C C O)
were prepared. Both series of com-
ꢀ
ꢁ
pounds, containing two reactive C CH
compact C O linked dendrimers
moieties on the dendrimer surface,
were used as macromonomers and co-
formed poly(dendrimer)s with a higher
degree of polymerization than the
polymerized with trans-[Pt
to form organometallic poly(dendri-
ACHTUNGTRENNUNGmer)s by an outer-sphere–outer-sphere
ACHTUGNRTNE(NUNG PEt3)2Cl2]
Keywords: dendrimers · platinum ·
polymerization · polymers
connection strategy. It was found that
Introduction
the first two classes of compounds, reports on the synthesis
and properties of poly(dendrimer)s are rare. Dendrimer–
linear polymer hybrids can be conveniently accessed syn-
thetically by attaching multiple linear polymer chains to the
surface functionalities or one linear polymer chain to the
focal point of a dendron. In contrast, dendronized polymers
may be prepared through the graft-to, graft-from, or macro-
monomer polymerization strategies.[2] Invariably, all three
methods involve reactions at the focal point functional
group (i.e. inner-sphere–inner-sphere connection)[4] of a den-
dron. As a result, steric inhibition could become a major
synthetic problem, especially in the formation of dendron-
ized polymers bearing higher generation dendritic side
chains. Some time ago we reported the direct 1:1 copolymer-
The use of dendrons or dendrimers as building blocks for
the synthesis of more complex polymer architectures is a
topic of current interest. Several classes of novel structures,
namely dendrimer–linear polymer hybrids,[1] dendronized
polymers,[2] and poly(dendrimer)s,[3] have appeared
(Figure 1). Although there are numerous examples based on
ization of trans-[PtACHTUNRGTNEUNG(PEt3)2Cl2] and dendrimers 1 bearing two
reactive functional groups on the surface of an oligoether
dendrimer to form poly(dendrimer)s 2[5] and anticipated
that steric hindrance could be alleviated by using an outer-
sphere–outer-sphere connection strategy.[4] Because of their
structural resemblance to necklaces, they were also called
dendritic necklaces. In this particular study, we deliberately
ꢀ
installed the reactive C CH functionality at the end of an
elongated C-10 surface functionality with the aim to im-
prove the polymerization efficiencies. As it turned out, the
degree of polymerization (DP) value was 103 (based on
laser light scattering data) for first generation (G1), 102 for
second generation (G2), and 30 for third generation (G3)
bifunctional dendrimers (Figure 2).
Figure 1. Schematic drawings of a) dendrimer–linear polymer hybrids,
b) a dendronized polymer, and c) a poly(dendrimer).
[a] S.-Y. Cheung, Prof. Dr. H.-F. Chow, Prof. Dr. T. Ngai, X. Wei
Department of Chemistry, The Chinese University of Hong Kong
Shatin, NT, Hong Kong SAR (China)
Fax : (+852)2603-5057
E-mail: hfchow@cuhk.edu.hk
[b] Prof. Dr. H.-F. Chow
The Center of Novel Functional Molecules
The Chinese University of Hong Kong (China)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200802306.
Figure 2. Synthesis of poly(dendrimer)s 2 from surface bifunctional mac-
romonomers 1.
Chem. Eur. J. 2009, 15, 2278 – 2288
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2279

H.-F. Chow et al.
To better evaluate this new approach for the synthesis of
poly(dendrimer)s, it is necessary to carry out a thorough in-
vestigation of how the various structural factors can affect
the polymerization efficiency. Herein, we report 1) the con-
trolled syntheses of two different series of G1–G3 surface-
bifunctional oligoether dendrimers, one containing a shorter
ence of 1% of monofunctional or trifunctional impurities is
not easily detectable by H NMR spectroscopy or size-exclu-
1
sion chromatography (SEC) because of their structural and
size resemblance, and hence one cannot rely only on spec-
troscopic means to confirm their purity. Rather, it is abso-
lutely necessary to design a synthetic route that can guaran-
tee the formation of the target bifunctional macromonomers
in extremely pure form.
ꢁ
C O branch skeleton S-Gn (n=1–3), and the other L-Gn
The retrosynthetic analysis of the bifunctional macro-
AHCTUNGTREGmNNUN onomers L-Gn and S-Gn is shown in Figure 3. Apart from
ꢁ ꢁ ꢁ
(n=1–3) containing a longer C C C O branch, in high
yields and excellent purity, 2) their 1:1 copolymerizations
ꢀ
with the linker trans-[Pt
G
CH units, and 3) the dependence of the polymerization effi-
ciency on the structural flexibility and molecular size of the
dendritic macromonomers. A comparison of the polymeri-
zation results of these two series of dendrimers, which differ
only by the length of the branches, is made to resolve the
issue of whether the structural rigidity of the branches may
have an effect on polymerization. It was also of interest to
see whether the polymerization of such large macromono-
mers with two reactive groups located at distant locations
has any different characteristics from those of conventional
small monomer molecules.
Figure 3. Retrosynthetic analysis of surface bifunctional macromonomers
L-Gn and S-Gn.
the final coupling step of a dendron 3 to the central core to
produce the target compounds, all other operations involve
the differential connection of a branching unit 6 to two dif-
ferent dendrons, that is, one symmetrical dendron (e.g. 4)
and one unsymmetrical dendron (e.g. 5). To avoid install-
ment of two identical dendrons to the branching unit 6, one
must exert complete control on the selective coupling of the
branching unit 6 to two different dendrons. This can be ac-
complished either by controlling the reactant stoichiometry
of the coupling reaction or by creating an appropriately pro-
tected branching unit. In addition, as the terminal acetylene
functionality is relatively labile, we plan to install it by So-
nogashira coupling[6] to the corresponding aryl iodide at the
postdendrimerization stage. Hence, our initial objective was
to prepare the two different series of surface diiodo-func-
Results and Discussion
Retrosynthetic analysis: The specific installment of only two
reactive surface groups on a dendrimer surface containing
multiple surface functionalities is by no means an easy task.
There are many complicating factors that one should consid-
er before carrying out the synthesis. First, only two, not one
or three, reactive groups should be decorated on the surface.
Second, the purity of the bifunctional dendrimers should be
extremely high. If, as a result of some synthetic flaws, 1% of
a monofunctional dendrimer is formed as a by-product
mixed with 99% of the bifunctional dendrimer, this mono-
functional dendrimer will then serve as a polymer chain
stopper, and the theoretical maximum DP value will
become 200. Similarly, if the macromonomer sample con-
tains 1% of a trifunctional macromonomer, a network struc-
ture will be formed after polymerization. Third, the pres-
ꢁ
ꢁ
ꢁ
ꢁ
tionalized dendrimers I ACHTUNRGTENN[UG S-Gn] I and I CAHTUNGTRENN[UGN L-Gn] I.
Synthesis of S-Gn dendrimers: The starting material for the
ꢁ
short (C O) branching unit was ethyl 3,5-dibenzyloxyben-
zoate 7 (Scheme 1). Controlled partial hydrogenolysis of
compound 7 in the presence of 10% Pd on C in EtOAc/
EtOH (1:1) at 208C afforded the monobenzylated branching
unit 8 in 52% yield. The fully deprotected ethyl 3,5-dihy-
droxybenzoate and the unreacted starting material 7 could
2280
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2278 – 2288

Synthesis of Organometallic Poly(dendrimer)s
FULL PAPER
15 in two steps as described ear-
lier for compound 9. The bro-
mide 15 was then anchored to
the hydroquinone core to pro-
duce the bifunctional dendri-
ꢁ
ꢁ
mer I ACHTNUTRGNEUNG[S-G2] I in 95% yield
under Williamson ether synthe-
sis conditions. Finally, the two
iodo functional groups were
similarly treated to give the
target G2 dendrimer S-G2 by
palladium catalyzed coupling
ꢀ
with TMSC CH followed by
deprotection of the TMS
groups with K2CO3 in MeOH in
91% overall yield.
Scheme 1. Synthesis of S-G1. Reagents and conditions: a) H2, 10% Pd-C, EtOAc/EtOH 1:1, 258C, 2 h, 52%;
b) 4-iodobenzyl bromide, K2CO3, [18]crown-6, acetone, 568C, 12 h, 97%; c) DIBAL-H, hexane, ꢁ608C!08C,
2 h, 52%; d) PPh3, CBr4, THF, 258C, 2 h, 61%; e) hydroquinone (0.45 equiv), Cs2CO3, dibenzo-[24]crown-8,
ꢀ
DMF, 258C, 12 h, 85%; f) TMSC CH, CuI, [PdACTHUNRTGNENG(U PPh3)2Cl2], PPh3, NEt3, toluene, 1008C, 3 days, 97%;
g) K2CO3, MeOH/THF 1:1, 258C, 2 h, 98%.
Preparation of the S-G3 den-
drimer
involved
selective
be removed cleanly by column chromatography because of
the large differences of their chromatographic mobilities.
The complete removal of ethyl 3,5-dihydroxybenzoate and
the unreacted starting material was of utmost importance
here. The presence of a tiny amount of such contaminants
carried through the rest of the synthesis would result in the
formation of an inseparable monofunctional polymer chain
stopper, a trifunctional or even a tetrafunctional dendrimer
cross-linker. Treatment of compound 8 with 4-iodobenzyl
bromide (1.1 equiv) in the presence of K2CO3 and
[18]crown-6 in refluxing acetone afforded the unsymmetrical
dendron 9 in 97% yield. No C-alkylation products were
found under the reaction conditions. The focal point ethyl
ester functionality was then transformed into the corre-
sponding benzyl bromide 10 in two steps by diisobutylalumi-
nium hydride (DIBAL-H) reduction and bromination (PPh3,
CBr4) in 59% yield. Coupling of 2.2 equiv of the bromide 10
with the central hydroquinone core furnished the bifunction-
Scheme 2. Synthesis of S-G2. Reagents and conditions: a) 12 (0.25 equiv),
K2CO3, [18]crown-6, acetone, 568C, 12 h, 79%; b) 10 (1.1 equiv), K2CO3,
[18]crown-6, acetone, 568C, 12 h, 91%; c) DIBAL-H, hexane, ꢁ608C!
08C, 2 h, 91%; d) PPh3, CBr4, THF, 258C, 2 h, 87%; e) hydroquinone
ꢁ
ꢁ
al G1 dendrimer I ACHTUNGTRENNUNG[S-G1] I in 85% yield under Williamson
ether synthesis conditions. The two iodo functional groups
ꢀ
(0.45 equiv), K2CO3, [18]crown-6, acetone, 568C, 3 days, 95%; f) TMSC
CH, CuI, [Pd
A
g) K2CO3, MeOH/THF 1:1, 258C, 2 h, 97%.
ꢀ
were then coupled to TMSC CH (TMS=trimethylsilyl) to
ꢀ ꢁ
ꢁ ꢀ
give TMSC C
A
the TMS groups using K2CO3 in MeOH to produce S-G1 in
95% overall yield.
mono-O-alkylation (K2CO3, [18]crown-6) of ethyl 3,5-dihy-
droxybenzoate 11 with the known G2-bromide 16
(0.25 equiv)[7] to give the hemisubstituted derivative 17 in
64% yield, free of the di-O-alkylation product after chroma-
tography purification (Scheme 3). A second O-alkylation
step (K2CO3, [18]crown-6) with the functionalized G2-Br 15
then produced the functionalized G3-CO2Et dendron 18 in
99% yield. The focal point ester moiety was then converted
to afford G3-benzyl bromide 19 derivative in 77% yield by
DIBAL-H reduction followed by bromination (CBr4, PPh3).
Compound 19 (2.2 equiv) was then secured onto the hydro-
Synthesis of the S-G2 dendrimer began with the known
3,5-dibenzyloxybenzyl bromide 12[7] (Scheme 2). Selective
mono-O-alkylation of ethyl 3,5-dihydroxybenzoate 11 with
bromide 12 (0.25 equiv) afforded compound 13 in 79%
yield (based on compound 12). Again, it was absolutely nec-
essary to separate a small amount of the di-O-alkylation
product (ca. 10%) and the starting material 11 from the
target product 13. Fortunately, this was an easy task because
of the very different chromatographic mobilities of the three
compounds. Compound 13 was then subjected to a second
O-alkylation step with the unsymmetrical G1-dendritic bro-
mide 10 to furnish the monofunctionalized G2-dendron 14
under Williamson conditions in 91% yield. The ethyl ester
was then converted into the corresponding benzyl bromide
ꢁ
ꢁ
quinone core to afford the diiodo dendrimer I ACTHNUTRGENUG[N S-G3] I in
ꢁ
ꢁ
95% yield. Palladium-catalyzed coupling of I
[S-G3] I with
ꢀ
TMSC CH occurred uneventfully to produce the bis(trime-
ꢀ ꢁ
ꢁ ꢀ
thylsilylacetylene) dendrimer TMSC C
U
Chem. Eur. J. 2009, 15, 2278 – 2288
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2281

H.-F. Chow et al.
The synthetic routes for the elongated series of dendri-
AHCTUNGTRENNUNG
Scheme 3. Synthesis of S-G3. Reagents and conditions: a) 16 (0.25 equiv),
K2CO3, [18]crown-6, acetone, 568C, 12 h, 64%; b) 15 (1.1 equiv), K2CO3,
[18]crown-6, acetone, 568C, 12 h, 99%; c) DIBAL-H, hexane, ꢁ608C!
08C, 2 h, 99%; d) PPh3, CBr4, THF, 258C, 2 h, 78%; e) hydroquinone
ꢀ
(0.45 equiv), K2CO3, [18]crown-6, acetone, 568C, 3 days, 95%; f) TMSC
CH, CuI, [Pd(PPh3)2Cl2], PPh3, NEt3, toluene, 1008C,
3
d, 86%;
24 (13%), and C,O-alkylated 25 (ca. 5%) products. Fortu-
nately, these products could be separated by column chro-
matography.
g) TBAF, THF, 258C, 15 min, 49%.
Structural characterization: The structures of all compounds
1
in 86% yield. However, removal of the TMS groups in the
presence of K2CO3 in MeOH (258C, 2 h) gave a mixture of
the target bifunctional dendrimer S-G3 together with 10%
of some oligomeric Hay homocoupling products[8] as sug-
were characterized by H and 13C NMR spectroscopy. The
1H NMR signals were well separated into different regions
and therefore spectral analyses were straightforward. For
the S-Gn series of compounds, the aromatic signals due to
the central hydroquinone core, the brancher, and the surface
groups were found at d=6.9, 6.7–6.4, and 7.5–7.2 ppm, re-
spectively, whereas the benzyl protons originating from the
Frꢁchet-type dendritic branches were found at d=5.0–
4.8 ppm. For the L-Gn series, the presence of the two extra
methylene groups on the dendritic branches gave two addi-
tional sets of aliphatic proton signals at d=2.8–2.6 and 2.2–
1.9 ppm, whereas those of the benzyl protons were located
at slightly higher field at d=4.0–3.8 ppm. The remaining
sets of aromatic signals were very similar to those of the S-
Gn series. Hence, the aromatic signals due to the central hy-
droquinone core, the brancher, and the surface groups were
found at d=6.8, 6.4–6.2, and 7.4–7.1 ppm, respectively. De-
pending on the nature of the functionalized surface group, a
1
gested by H NMR spectroscopy and SEC analysis. We sus-
pected that the occurrence of Hay homocoupling was due to
a contamination of trace amount of copper salts left over
from the previous reaction. Fortuitously, this side reaction
could be totally suppressed if tetrabutylammonium fluoride
(TBAF) was used (208C, 15 min) as the deprotecting agent.
However, homocoupling products could still be formed if
the reaction was left to run for longer than 15 min.
Synthesis of L-Gn dendrimers: Three key intermediates
were required for the synthesis of the elongated bifunctional
dendrimers L-Gn. One was the surface-functionalized iodi-
nated derivative 20, the second was the commercially avail-
able unfunctionalized surface derivative-1-bromo-3-phenyl-
propane 21, and the third was the elongated branching
agent 22.[5] The surface unit 20 could be conveniently pre-
pared from hydrocinnamic acid by using an improved
method described in the Supporting Information.
ꢀ
sharp signal at about d=0.2 ppm was found for the TMSC
C functionalized compounds, whereas the acetylenic proton
signal of the HC C containing compounds was found at
ꢀ
around d=3.0 ppm.
2282
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2278 – 2288

Synthesis of Organometallic Poly(dendrimer)s
FULL PAPER
The 13C NMR spectral features of the synthesized com-
pounds were also consistent with the proposed structures.
For the S-Gn series, the aromatic C signals of the surface,
branching, and central core were found to spread within a
wide window of d=160–100 ppm, whereas those of the ben-
zylic C atoms were clustered at d=71–69 ppm. For the iodo
surface-functionalized compounds, the signal of the aromatic
C atoms directly attached to the iodine atom was found at
d=94 ppm, while two extra acetylenic carbon signals were
the target compounds. For all the iodo-functionalized den-
drons, the major signals are the molecular peaks due to
[M]+, [M+H]+, [M+Na]+ or [M+K]+. For example, the MS
spectrum of the mono-iodinated G3 dendron 19 gave a base
peak signal corresponding to [M+H]+ at m/z 1783.4
(Figure 5). The other minor signals are due to fragmentation
ꢀ
ꢀ
found for the TMSC C (d=105 and 95 ppm) and HC C
(d=84 and 77 ppm) surface-functionalized compounds. Sim-
ilarly trends were also observed for the L-Gn series, except
for the presence of aliphatic carbon signals (d=33–30 ppm)
belonging to the longer branches.
The purities of the compounds were determined by SEC
using polystyrenes as the standards. All compounds gave a
sigmoidal peak with a polydispersity index (PDI) of less
than 1.03. As expected, the retention times of the fully alky-
lated dendrons having the same generation number were
very close to each other, whereas the bifunctional dendri-
ACHTUNGTRENNUNGmers of the same generation possessed slightly shorter re-
tention times. Hence, in the SEC scatter plot of retention
time versus theoretical molecular weight (MW), dendrons of
the same generation number clustered around each other
(Figure 4). Furthermore, the shorter series (S) of compounds
Figure 5. Mass spectrum of mono-iodinated G3 dendron 19, showing the
M+H+ peak at m/z 1783. The M+ peak of the de-iodinated species has
an m/z value of 1657.
ions. However, we could not find the signal due to the corre-
sponding de-iodinated dendron that could potentially be
formed during DIBAL-H reduction of the corresponding
G3 ester dendron 18.
Copolymerization studies: Having obtained the target bi-
functional dendrimers, we decided to find out the optimized
conditions to carry out their 1:1 copolymerizations with the
platinum linker trans-[PtACHTUNGTRENUNG(PEt3)2Cl2] in the presence of CuI
and diisopropylamine in CHCl3 at 408C. With S-G1 as the
model substrate, the copolymerization was conducted under
different monomer concentrations of 4.4, 8.7, and 17 mm. It
was found that at low concentration (4.4 mm), the product
contained a significant amount of low molecular weight
(LMW) cyclic oligomers (see below) as determined by SEC
and MS analyses (see the Supporting Information for de-
tails). The amount of LMW oligomers was much less when
the polymerizations were conducted at 8.7 and 17 mm. How-
ever, owing to the much larger MW of the G3 monomers,
their copolymerizations at 17 mm resulted in a very viscous
solution, which significantly reduced the stirring efficiency.
Hence, all copolymerizations were carried out at 8.7 mm at
408C under a closed system to prevent solvent evaporation.
The copolymerizations of the two series of bifunctional
dendrimers were then conducted side by side under identical
reaction conditions to evaluate the structural effect on the
formation of poly(dendrimer)s (Scheme 4). After polymeri-
zation, the yellow products Pt-S-Gn and Pt-L-Gn were iso-
lated by passing them through a short column of alumina
followed by solvent evaporation, and they were then sub-
Figure 4. Theoretical MW versus SEC retention time cluster plot of den-
drons of various generations.
possessed a longer retention time than the corresponding
analogue in the elongated branch series (L), indicating that
the elongated series of compounds have a slightly larger hy-
drodynamic radius.
It was mentioned earlier that the presence of a small
amount of monofunctional dendrimer could have a detri-
mental effect on the polymerization efficiency. However, the
nearly 100% structural purities of these compounds could
not be assessed from the relative integration values of the
signals for the functionalized surface versus the unfunction-
1
alized surface groups by H NMR spectroscopy. Mass spec-
troscopic (MS) analysis, in particular FAB/MALDI-TOF,
may provide a better estimate of the structural purities of
Chem. Eur. J. 2009, 15, 2278 – 2288
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2283

H.-F. Chow et al.
Scheme 4. Copolymerizations of S-Gn and L-Gn with trans-[Pt-
ACHTUNGTRENNUNG(PEt3)2Cl2]. Reagents and conditions: a) CuI, HNPr2, CHCl3 , 408C, 72 h.
1
jected to H, 13C, and 31P NMR spectroscopic characteriza-
tion. For example, the 1H NMR spectrum of Pt-S-G2 re-
vealed the complete disappearance of the acetylenic proton
signals (d=3.07 ppm) and the upfield shifting of the aromat-
ic signals ortho to the functional 4-ethynylphenyl surface
moieties (Figure 6). Furthermore, the signals due to the tri-
Figure 7. 31P{1H} NMR spectra (122 MHz, CDCl3) of Pt-S-G2 (top),
showing two sets of 31P signals (square and eclipse), and Pt-S-G3
(bottom), showing three sets of 31P signals (square, eclipse, and triangle).
Table 1. SEC (408C) and LLS data (258C) of poly(dendrimer)s in THF.
Sample
SEC Mw ꢂ10ꢁ3 DP[a] PDI[a] % of LMW LLS Mw ꢂ10ꢁ3
oligomers[a,b]
[c]
Pt-S-G1
Pt-L-G1
Pt-S-G2 122
Pt-L-G2
Pt-S-G3
Pt-L-G3
40
53
34
39
60
22
21
19
1.9
1.9
2.3
1.6
1.8
1.6
1
2
–
–
[c]
<0.2
4
3
71
[c]
55
79
86
–
64
[c]
14
–
[a] Values calculated from SEC data; [b] wt% up to the pentamer;
[c] Mw value of the non-aggregated poly(dendrimer) could not be deter-
mined accurately by LLS because of the relatively large intensity of the
aggregated peak.
Figure 6. 1H NMR spectra (300 MHz, CDCl3) of S-G2 (top) and Pt-S-G2
(bottom). The signal labeled with an asterisk is due to residual H2O.
values might deviate significantly from the actual figures.
Nonetheless, the conclusions drawn below regarding to the
relative polymerization efficiencies of these two series are
still valid. Because of the step-growth nature of the polymer-
ization, the polydispersity index (PDI) values were close to
2. However, several interesting findings are noteworthy.
First, a significant amount (0.2–14% by weight) of low mo-
lecular weight (LMW) oligomers (calculated up to the pen-
tamer) was found. Initially we thought that these were the
partially reacted open-chain oligomers still bearing either
ACHTUNGTRENNUNGethylphosphine ligands (d=2.25–2.10 and 1.26–1.10 ppm) at-
tached to the platinum metal could also be found. The
31P{1H} NMR spectra exhibited one major 31P peak located
at d=11.1 ppm with two 195Pt satellite signals [1J
ACTHNUTRGNE(NUG Pt,P)
ꢂ2360 Hz] of one sixth intensity (Figure 7). For the higher
generation G3 poly(dendrimer)s Pt-S-G3 and Pt-L-G3,
minor peaks at d=8.6 and 14.9 ppm were also observed,
suggesting that the polymerizations were of lower efficiency.
The molecular weights (Mw) and distributions of the poly-
ꢁ
(dendrimer)s were determined by SEC and static and dy-
free acetylene or Pt Cl chain ends. However, subjecting the
namic laser light scattering (LLS) studies (Table 1 and
Figure 8). Because of the formation of poly(dendrimer) ag-
gregates in concentrated THF and also in concentrated tolu-
ene solutions, their Mw values, apart from for Pt-S-G2 and
Pt-S-G3, could not be accurately determined by LLS (see
then Supporting Information for details). Hence, the tabu-
lated DP values are based on the Mw value obtained from
SEC calculations. Aggregate formation was not found in
SEC measurements as they were carried out with highly di-
luted solutions in THF. It should also be noted that polysty-
polymer mixture to further reaction did not result in a de-
crease in the signal intensities of such LMW oligomeric
peaks. Most interestingly, apart from the SEC trace of the
products from S-G1, the rest of the SEC traces showed the
presence of a peak (labeled with an asterisk in Figure 8)
that possessed a longer retention time than that of the cor-
responding dendritic macromonomer. Hence, such peaks
most likely arose from the cyclic monomer. Indeed, this was
confirmed by the MS data of the purified samples by prepa-
rative SEC. For example, for the SEC trace of Pt-L-G1, the
signal peak with retention time of 33 min had a molecular
ACHTUNGTRENNUNGrenes were used as standards and hence the calculated Mw
2284
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2278 – 2288

Synthesis of Organometallic Poly(dendrimer)s
FULL PAPER
L-Gn series, suggesting that more cyclic oligomers were
formed from the L-Gn series. This difference was more pro-
nounced for the G3 compounds, for which 3% of LMW
oligomers was found for the S series, in contrast to 14% for
the L series. This finding revealed a hitherto unknown
factor in the polymerization of macromonomers with two
remote reactive functionalities separated at a distance,
namely, that structurally more flexible monomers are more
prone to undergo cyclization than linear propagation rela-
tive to their rigid analogues.
Our third observation was that the polymerization effi-
ciency within the same series of dendrimers was strongly de-
pendent on the dendrimer generation. Both the G1 and G2
dendrimers polymerized with a larger DP value than the G3
dendrimers. The drop of the DP values for the G3 poly(den-
drimer)s could be attributed to the increased steric inhibi-
tion during polymerization, and also to the lowering of sur-
face density of the acetylenic moiety (from 1/2 of G1 den-
drimers to 1/8 of the G3 dendrimers). A comparison of the
SEC calculated DP values (19–60) in the current study to
those (20–63 based on SEC data) obtained in our previous
study[5] involving elongated C-10 bifunctional macromono-
ꢀ
mers 1 suggested that having a protruding C CH reactive
functionality sticking up from the dendrimer surface did not
possess any obvious advantage in terms of polymerization
efficiency. As the SEC calculated Mw values were known to
be grossly underestimated by polystyrene standards,[9] the
actual DP values of such poly(dendrimer)s could be very
high. Unfortunately, we were unable to extract reliable Mw
data from static or dynamic LLS studies to support this
claim.
Conclusion
Figure 8. SEC chromatograms of a) Pt-S-G1, b) Pt-L-G1, c) Pt-S-G2,
d) Pt-L-G2, e) Pt-S-G3, and f) Pt-L-G3 (solid lines). The peaks labeled
with an asterisk are cyclic monomer peaks. The dotted lines are the SEC
traces of the corresponding dendritic macromonomers. The wt% of the
oligomers is exaggerated in these plots as the x axis is on a logarithmic
scale of polymer MW.
In summary, we report herein the controlled syntheses of
surface bifunctional dendrimers S-Gn and L-Gn by a con-
vergent synthetic strategy. Such dendritic macromonomers
were used to copolymerize with trans-[PtACHTNUGTRNEG(UN PEt3)2Cl2] to pro-
duce the corresponding poly(dendrimer)s Pt-S-Gn and Pt-
L-Gn, respectively. The polymerization efficiency was found
to depend on the monomer concentration. In general, lower
monomer concentration led to the formation of more cyclic
oligomers. It was also revealed that the structurally more
rigid S-Gn series underwent polymer propagation to form
linear poly(dendrimer)s much more effectively. In contrast,
the structurally more flexible L-Gn series had a relatively
higher tendency to form cyclic oligomers. Furthermore, the
presumed advantage of protruding the reactive functional
group out of the dendrimer surface to improve the polymer-
ization efficiency could not be substantiated in this investi-
gation. Our studies revealed that there are additional struc-
tural factors that are not encountered in the polymerization
of conventional small monomers are involved in the poly-
merization of such macromonomers.
peak at m/z 1361.6254, which corresponded to the structure
of the cyclic monomer 26 (see the Supporting Information
for MS data of cyclic monomers of other poly(dendrimer)s).
The second point worth mentioning is that the amount of
LMW oligomers of the S-Gn series was less than that of the
Chem. Eur. J. 2009, 15, 2278 – 2288
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2285

H.-F. Chow et al.
1H NMR (300 MHz, CDCl3): d=7.47 (d, 3J
7.28–7.44 (m, 14H; ArH), 6.87 (s, 4H; core-ArH), 6.68 (s, 2H; ArH),
6.65 (s, 2H; ArH), 6.53 (t, 3J
(H,H)=2.1 Hz, 2H; ArH), 5.03 (s, 4H;
ArCH2O), 5.02 (s, 4H; ArCH2O), 4.95 (s, 4H; ArCH2O), 0.25 ppm (s,
18H; Si
(CH3)3); 13C NMR (75 MHz, CDCl3): d=160.2, 160.0, 153.2,
139.9, 137.4, 136.8, 132.3, 128.7, 128.1, 127.7, 127.2, 122.8, 115.9, 106.5,
106.4, 104.9, 101.6, 94.6, 70.6, 70.2, 69.7, 0.1 ppm; MS (ESI): m/z (%): 929
(100) [M+Na+]; HRMS (ESI) calcd for C58H58O6Si2+Na+: 929.3664;
found: 929.3680; elemental analysis calcd (%) for C58H58O6Si2: C 76.78,
H 6.44; found: C 76.43, H 6.11, N <0.10.
ACHTUNGERTN(NUNG H,H)=8.1 Hz, 4H; ArH),
Experimental Section
AHCTUNGTRENNUNG
ꢁ
ꢁ
I ACHTUNGTRENNUNG[S-G1] I: A mixture of hydroquinone (0.803 g, 7.29 mmol), bromide
10 (8.17 g, 16.0 mmol), Cs2CO3 (8.32 g, 25.5 mmol), and dibenzo-
[24]crown-8 (5 mg) in DMF (40 mL) was stirred under N2 at 258C for
12 h. The reaction mixture was filtered and washed with CH2Cl2. The fil-
trates were combined, dried in vacuo, and purified by flash chromatogra-
phy (eluent: hexane/CHCl3 1:1) to give the target compound (5.98 g,
85%) as a white solid. M.p. 136.5–137.58C; Rf =0.30 (hexane/CHCl3 1:1);
AHCTUNGTRENNUNG
1H NMR (300 MHz, CDCl3): d=7.71 (d, 3J
7.29–7.48 (m, 10H; ArH), 7.16 (d, 3J
(H,H)=8.4 Hz, 4H; ArH), 6.88 (s,
4H; core-ArH), 6.70 (s, 2H; ArH), 6.66 (s, 2H; ArH), 6.54 (t, J
ACHTUNGERTN(NUNG H,H)=8.4 Hz, 4H; ArH),
ꢀ ꢁ
[S-G2]
ꢁ ꢀ
CTMS: Starting from
ꢁ
ꢁ
TMSC
C
G
C
I ACTHUNGTERNNU[G S-G2] I (408 mg,
AHCTUNGTRENNUNG
3
0.22 mmol), [(PPh3)2PdCl2] (79 mg, 0.11 mmol), PPh3 (29 mg, 0.11 mmol),
CuI (21 mg, 0.11 mmol), trimethylsilylacetylene (0.48 mL, 3.37 mmol),
and Et3N (0.47 mL, 3.37 mmol) in toluene (10 mL), the product (369 mg,
94%) was obtained as a pale yellow foam after flash chromatography
(eluent: hexane/CH2Cl2 4:5 to CH2Cl2). Rf =0.54 (hexane/CH2Cl2 2:3);
AHCTUNGTRENGN(UN H,H)=
2.1 Hz, 2H; ArH), 5.04 (s, 4H; ArCH2O), 4.98 (s, 4H; ArCH2O),
4.96 ppm (s, 4H; ArCH2O); 13C NMR (75 MHz, CDCl3): d=160.2, 159.9,
153.1, 139.9, 137.7, 136.8, 136.6, 129.4, 128.7, 128.1, 127.6, 115.9, 106.5,
106.3, 101.5, 93.6, 70.5, 70.1, 69.4 ppm; MS (EI): m/z (%): 966 (<1) [M+
]; HRMS (EI) calcd for C48H40I2O6+: 966.0909; found: 966.0913.
1H NMR (300 MHz, CDCl3): d=7.47 (d, 3J
7.27–7.45 (m, 34H; ArH), 6.88 (s, 4H; core-ArH), 6.68–6.72 (m, 6H;
ArH), 6.61–6.68 (m, 6H; ArH), 6.58 (t, 3J
(H,H)=2.1 Hz, 2H; ArH),
ACHTUNGERTN(NUNG H,H)=7.8 Hz, 4H; ArH),
ꢁ
ꢁ
I ACHTUNGTRENNUNG[S-G2] I: A mixture of hydroquinone (0.16 g, 1.46 mmol), bromide 14
AHCTUNGTRENNUNG
(3.00 g, 3.21 mmol), K2CO3 (0.61 g, 4.38 mmol), and [18]crown-6 (2 mg)
in acetone (80 mL) was heated to reflux for 3 days. The mixture was fil-
tered and washed with CH2Cl2. The combined filtrates were evaporated
in vacuo and the residue was chromatographed on silica gel to give the
product (2.52 g, 95%) as a white foam (eluent: hexane/CH2Cl2 1:2 to
6.50–6.56 (m, 4H; ArH), 5.03 (s, 12H; PhCH2O), 5.02 (s, 4H; ArCH2O),
4.98 (s, 4H; ArCH2O), 4.97 (s, 4H; ArCH2O), 4.93 (s, 4H; ArCH2O),
0.27 ppm (s, 18H; SiACTHNUTRGNEUNG
(CH3)3); 13C NMR (75 MHz, CDCl3): d=160.1,
160.0, 159.9, 153.1, 139.8, 139.3, 137.3, 136.8, 132.1, 128.6, 128.0, 127.6,
127.2, 122.7, 115.8, 106.4, 105.0, 101.6, 94.5, 70.4, 70.0, 69.9, 69.5, 0.1 ppm;
MS (ESI): m/z (%): 1756 (75) [M+H+]; HRMS (ESI) calcd for
C114H106O14Si2+H+: 1755.7194; found: 1755.7201; elemental analysis calcd
(%) for C114H106O14Si2: C 77.96, H 6.08; found: C 77.69, H 5.96, N <0.10.
1
CH2Cl2). Rf =0.33 (hexane/CH2Cl2 1:2); H NMR (300 MHz, CDCl3): d=
7.68 (d, 3J
3J
ArH), 6.60–6.66 (m, 6H; ArH), 6.57 (t, J
(t, 3J
(H,H)=2.1 Hz, 4H; ArH), 5.03 (s, 12H; PhCH2O), 4.97 (s, 4H;
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
3
AHCTUNGTERG(NNUN H,H)=2.1 Hz, 2H; ArH), 6.53
ꢀ ꢁ
[S-G3]
ꢁ ꢀ
C CTMS: Starting from
ꢁ
ꢁ
TMSC
C
G
I ACTHUNGTERNNU[G S-G3] I (346 mg,
ACHTUNGTRENNUNG
ArCH2O), 4.96 (s, 8H; ArCH2O), 4.92 ppm (s, 4H; ArCH2O); 13C NMR
(75 MHz, CDCl3): d=160.1, 160.0, 159.9, 159.8, 153.0 ,139.8, 139.3, 139.2,
137.5, 136.73, 136.68, 136.4, 129.3, 128.5, 128.0, 127.5, 115.7, 106.4, 106.3,
101.5, 93.5, 70.3, 69.9, 69.83, 69.77, 69.2 ppm; MS (ESI): m/z (%): 1816
(100) [M+H+]; HRMS (ESI) calcd for C104H88I2O14+H+: 1815.4336;
found: 1815.4327; elemental analysis calcd (%) for C104H88I2O14: C 68.80,
H 4.89; found: C 68.76, H 4.56, N <0.10.
0.098 mmol), [(PPh3)2PdCl2] (79 mg, 0.049 mmol), PPh3 (29 mg,
0.049 mmol), CuI (21 mg, 0.049 mmol), trimethylsilylacetylene (0.48 mL,
1.48 mmol), and Et3N (0.47 mL, 1.48 mmol) in toluene (3.5 mL), the
product (294 mg, 86%) was obtained as a pale yellow foam after flash
chromatography (eluent: hexane/CH2Cl2 1:2 to CH2Cl2 to CH2Cl2/Et2O
100:1). Rf =0.46 (hexane/CH2Cl2 1:2); 1H NMR (300 MHz, CDCl3): d=
7.45 (d, 3J
4H; core-ArH), 6.59–6.71 (m, 28H; ArH), 6.55 (t, 3J
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
ꢁ
ꢁ
I ACHTUNGTRENNUNG[S-G3] I: A mixture of hydroquinone (13.6 mg, 0.124 mmol), bromide
ArH), 6.48–6.54 (m, 6H; ArH), 4.99 (s, 28H; PhCH2O), 4.98 (s, 4H;
ArCH2O), 4.94 (s, 24H; ArCH2O), 4.88 (s, 4H; ArCH2O), 0.25 ppm (s,
19 (484 mg, 0.272 mmol), K2CO3 (51.2 mg, 0.371 mmol), and [18]crown-6
(2 mg) in acetone (10 mL) was heated to reflux for 3 days. The mixture
was filtered and washed with CH2Cl2. The combined filtrates were con-
centrated in vacuo and the residue was chromatographed on silica gel to
give the product (412 mg, 95%) as a colorless oil (eluent: hexane/CH2Cl2
1:1 to CH2Cl2 to CH2Cl2/Et2O 100:1). Rf =0.52 (hexane/CH2Cl2 1:2);
18H; SiACHTUNGTRNEUNG
(CH3)3); 13C NMR (75 MHz, CDCl3): d=160.2, 160.14, 160.05,
160.0, 159.9, 153.1, 139.8, 139.32, 139.26, 137.3, 136.82, 136.79, 136.76,
132.1, 128.6, 128.0, 127.6, 127.2, 122.7, 115.8, 106.4, 105.0, 101.6, 94.5,
70.4, 70.0, 69.9, 69.5, 0.0 ppm; HRMS (MALDI-TOF) calcd for
C226H202O30Si2+Na+: 3476.3774; found: 3476.3867; elemental analysis
calcd (%) for C226H202O30Si2: C 78.58, H 5.89; found: C 78.37, H 5.83, N
<0.10.
1H NMR (300 MHz, CDCl3): d=7.67 (d, 3J
7.27–7.48 (m, 70H; ArH), 7.12 (d, 3J
(H,H)=8.1 Hz, 4H; ArH), 6.85 (s,
ACHTUNGERTN(NUNG H,H)=8.1 Hz, 4H; ArH),
AHCTUNGTRENNUNG
4H; core-ArH), 6.61–6.74 (m, 28H; ArH), 6.49–6.61 (m, 14H; ArH),
5.01 (s, 28H; PhCH2O), 4.95 (s, 24H; ArCH2O), 4.93 (s, 4H; ArCH2O),
4.89 ppm (s, 4H; ArCH2O); 13C NMR (75 MHz, CDCl3): d=160.2,
160.14, 160.11, 160.09, 159.9, 153.2, 139.9, 139.41, 139.39, 139.36, 139.3,
137.7, 136.9, 136.8, 136.6, 129.4, 128.7, 128.1, 128.0, 127.6, 115.9, 106.6,
106.5, 106.4, 101.7, 93.6, 70.5, 70.1, 70.0, 69.4 ppm; HRMS (MALDI-
TOF) calcd for C216H184I2O30+Na+: 3536.0921; found: 3536.1139; elemen-
tal analysis calcd (%) for C216H184I2O30: C 73.84, H 5.28; found: C 73.64,
H 5.21, N <0.10.
ꢀ ꢁ
ꢁ ꢀ
S-G1: A mixture of TMSC C [S-G1] C CTMS (320 mg, 0.35 mmol)
AHCTUNGTRENNUNG
and K2CO3 (240 mg, 1.76 mmol) in THF/MeOH (v/v 1:1, 20 mL) was
stirred at 258C for 2 h. The reaction mixture was filtered and washed
with CH2Cl2. The combined filtrates were dried in vacuo and purified by
flash column chromatography to give the target product (265 mg, 98%)
as a colorless oil after flash chromatography (eluent: hexane/CH2Cl2 1:1
to CH2Cl2 to CH2Cl2/Et2O 100:1). Rf =0.53 (hexane/CH2Cl2 1:4);
1H NMR (300 MHz, CDCl3): d=7.51 (d, 3J
7.29–7.46 (m, 14H; ArH), 6.87 (s, 4H; core-ArH), 6.69 (s, 2H; ArH),
6.66 (s, 2H; ArH), 6.55 (t, 3J
(H,H)=2.1 Hz, 2H; ArH), 5.04 (s, 8H;
ACHTUNGERTN(NUNG H,H)=8.1 Hz, 4H; ArH),
ꢀ ꢁ
[S-Gn]
ꢁ ꢀ
CTMS: A
General procedure for the synthesis of TMSC
C
G
C
ꢁ
ꢁ
ACHTUNGTRENNUNG
mixture of
I ACHTUNGTRENNUNG[S-Gn] I (1.0 equiv), [(PPh3)2PdCl2] (0.5 equiv), PPh3
PhCH2O+ArCH2O), 4.95 (s, 4H; ArCH2O), 3.09 ppm (s, 2H; C=CH);
13C NMR (75 MHz, CDCl3): d=160.3, 160.0, 153.1, 140.0, 137.7, 136.8,
132.5, 128.7, 128.2, 127.7, 127.4, 121.8, 115.9, 106.5, 106.4, 101.6, 83.5,
77.6, 70.6, 70.2, 69.6 ppm; MS (ESI): m/z (%): 763 (100) [M+H+];
HRMS (ESI) calcd for C52H42O6+H+: 763.3054; found: 763.3060; ele-
mental analysis calcd (%) for C52H42O6: C 81.87, H 5.55; found: C 82.25,
H 5.63, N <0.10.
(0.5 equiv), CuI (0.5 equiv), trimethylsilylacetylene (15 equiv), and Et3N
(15 equiv) in toluene was frozen in a sealed tube by liquid N2 and de-
gassed with N2 (3 times). The mixture was allowed to warm to 258C and
then heated at 1008C for 3 days. After the reaction was completed, Et2O
was added and the reaction mixture was filtered. The filtrate was dried in
vacuo and purified by flash column chromatography to give the target
compound.
ꢀ ꢁ
ꢁ ꢀ
ꢀ ꢁ
[S-G1]
ꢁ ꢀ
CTMS: Starting from
ꢁ
ꢁ
S-G2: A mixture of TMSC C [S-G2] C CTMS (701 mg, 0.40 mmol)
AHCTUNGTRENNUNG
TMSC
C
G
C
I ACTHUNGTERNNU[G S-G1] I (610 mg,
and K2CO3 (280 mg, 2.00 mmol) in THF/MeOH (v/v 1:1, 10 mL) was
stirred at 258C for 2 h. The reaction mixture was filtered and washed
with CH2Cl2. The combined filtrates were dried in vacuo and purified by
flash column chromatography to give the target compound (623 mg,
97%) as a white foam after flash chromatography (eluent: hexane/
0.63 mmol), [(PPh3)2PdCl2] (220 mg, 0.32 mmol), PPh3 (83 mg,
0.32 mmol), CuI (60 mg, 0.32 mmol), trimethylsilylacetylene (1.34 mL,
0.95 mmol), and Et3N (1.32 mL, 0.95 mmol) in toluene (20 mL), the prod-
uct (556 mg, 97%) was obtained as a pale yellow foam after flash chro-
matography (eluent: hexane/EtOAc 8:1). Rf =0.24 (hexane/EtOAc 8:1);
2286
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2278 – 2288

Synthesis of Organometallic Poly(dendrimer)s
FULL PAPER
CH2Cl2 1:2 to CH2Cl2 to CH2Cl2/Et2O 100:1). Rf =0.29 (hexane/CH2Cl2
tation) was obtained as a yellow powder. 1H NMR: (300 MHz, CDCl3):
d=7.15–7.47 (m, 78H; ArH), 6.84 (s, 4H; core-ArH), 6.66 (brs, 28H;
ArH), 6.55 (brs, 14H; ArH), 4.99 (s, 32H; ArCH2O), 4.94 (s, 24H;
ArCH2O), 4.88 (s, 4H; ArCH2O), 1.95–2.39 (m, 12H; PCH2), 1.05–
1.38 ppm (m, 18H; PCH2CH3); 13C NMR (75 MHz, CDCl3): d=160.2,
160.1, 153.1, 139.8, 139.31, 139.26, 139.2, 136.8, 133.3, 131.1, 128.6, 128.1,
1:2); 1H NMR (300 MHz, CDCl3): d=7.49 (d, 3J
ACHTUNGTRENNUNG
3
6H; ArH), 6.61–6.67 (m, 6H; ArH), 6.57 (t, JACTHNUTRGENUG(N H,H)=2.1 Hz, 2H; ArH),
6.49–6.56 (m, 4H; ArH), 5.02 (s, 16H; PhCH2O+ArCH2O), 4.97 (s, 8H;
ArCH2O), 4.92 (s, 4H; ArCH2O), 3.07 ppm (s, 2H; C CH); 13C NMR
ꢀ
ꢀ
ꢀ
127.6, 127.5, 115.8, 109.4 (C CPt), 108.4 (C CPt, weak signal), 106.4,
101.6, 70.5, 70.1, 70.0, 16.4 (quintet like, 1J
(C,P)=17.7 Hz; PCH2),
(Pt,P)=
(Pt,P)=
(75 MHz, CDCl3): d=160.2, 160.12, 160.08, 160.0, 153.2, 139.9, 139.4,
139.3, 137.7, 136.9, 136.8, 132.4, 128.7, 128.1, 127.7, 127.3, 121.8, 115.9,
106.6, 106.5, 106.4, 101.69, 101.66, 101.6, 83.5, 77.6, 70.6, 70.2, 70.04,
69.98, 69.6 ppm; MS (ESI): m/z (%): 1612 (100) [M+H+]; HRMS
(MALDI-TOF) calcd for C108H90O14+Na+: 1634.6257; found: 1634.6292;
elemental analysis calcd (%) for C108H90O14: C 80.48, H 5.63; found: C
80.10, H 5.30, N <0.10.
AHCTUNGTRENNUNG
8.4 ppm; 31P NMR (122 MHz, CDCl3): d=(major peak): 11.1 (1J
ACHTUNGTRENNUNG
2370 Hz); (minor peaks): 14.9 (1J(Pt,P)=2380 Hz), 8.6 ppm (1J
ACTHNUGTRENNUGN ACHTUNGTRENNUNG
2320 Hz).
Pt-L-G1: Starting from L-G1 (122 mg, 0.131 mmol), trans-[PtACHTUNGTRENNUNG(PEt3)2Cl2]
(65.8 mg, 0.131 mmol), and CuI (12.5 mg, 0.066 mmol) in CHCl3/iPr2NH
(v/v 1:1, 15 mL), the product (168 mg, 94%; 88 mg, 49% after precipita-
tion) was obtained as a yellow powder. 1H NMR (300 MHz, CDCl3): d=
7.12–7.35 (m, 14H; ArH), 6.98–7.12 (m, 4H; ArH), 6.90 (s, 4H, core-
ArH), 6.23–6.42 (m, 6H; ArH), 3.73–4.02 (m, 12H; ArOCH2); 2.53–2.94
(m, 12H; ArCH2+PhCH2), 2.11–2.43 (m, 12H; PCH2), 1.88–2.11 (m,
12H; ArCH2CH2+PhCH2CH2), 1.07–1.37 ppm (m, 18H; PCH2CH3);
13C NMR (75 MHz, CDCl3): d=160.3, 160.1, 153.3, 144.2, 144.0, 141.7,
ꢀ ꢁ
ꢁ ꢀ
S-G3: A mixture of TMSC C [S-G3] C CTMS (357 mg, 10.3 mmol)
AHCTUNGTRENNUNG
and TBAF (1.0m in THF, 0.3 mL) in THF (10 mL) was stirred at 258C
for 15 min. The reaction mixture was filtered and washed with CH2Cl2.
The combined filtrates were dried in vacuo and purified by flash column
chromatography to give the target compound (169 mg, 49%) as a pale
yellow foam after flash column chromatography (eluent: hexane/CH2Cl2
1:1 to CH2Cl2 to CH2Cl2/Et2O 50:1). Rf =0.39 (hexane/CH2Cl2 1:3);
1H NMR (300 MHz, CDCl3): d=7.46 (d, 3J
ACTHNUTRGNEUNG(H,H)=8.4 Hz, 4H; ArH),
7.27–7.43 (m, 74H; ArH), 6.82 (s, 4H; core-ArH), 6.59–6.73 (m, 28H;
ArH), 6.46–6.59 (m, 14H; ArH), 4.99 (s, 32H; PhCH2O+ArCH2O), 4.94
ꢀ
138.4, 131.0, 128.7, 128.5, 128.3, 126.6, 126.0, 115.5, 109.2 (C CPt), 107.7,
107.3, 107.2, 106.8 (t, 2J
A
ꢀ
1
32.6, 32.3, 32.1, 31.5, 31.3, 31.0, 30.9, 16.4 (quintet like, J
(C,P)=17.5 Hz;
PCH2), 8.5 ppm; 31P NMR (122 MHz, CDCl3): d=(major peak): 11.0 (1J-
(Pt,P)=2380 Hz); (minor peaks): 14.8 (t, 1J(Pt,P)=2390 Hz), 10.4 (1J-
(Pt,P)=2370 Hz), 8.5 ppm (1J
(Pt,P)=2330 Hz).
Pt-L-G2: Starting from L-G2 (263 mg, 0.131 mmol), trans-[Pt
ꢀ
(s, 24H; ArCH2O), 4.87 (s, 4H; ArCH2O), 3.06 ppm (s, 2H; C CH);
13
A
ACHTUNGTRENNUNG
ꢀ
C NMR (75 MHz, CDCl3, one of the C C signals was too weak to be
A
ACHTUNGTRENNUNG
observed): d=160.3, 160.2, 160.1, 160.0, 153.2, 139.9, 139.4, 139.3, 137.7,
136.9, 136.8, 132.4, 128.7, 128.1, 127.7, 127.4, 121.8, 115.9, 106.5, 101.7,
83.5, 70.6, 70.2, 70.1, 69.6 ppm; HRMS (MALDI-TOF) calcd for
C220H186O30+Na+: 3332.2988; found: 3332.2790; elemental analysis calcd
(%) for C220H186O30: C 79.83, H 5.66; found: C 79.77, H 5.66.
ACHTUNGTRENNUNG(PEt3)2Cl2]
(65.8 mg, 0.131 mmol), and CuI (12.5 mg, 0.066 mmol) in CHCl3/i-Pr2NH
(v/v 1:1, 15 mL), the product (286 mg, 90%; 227 mg, 71% after precipita-
tion) was obtained as a yellow powder. 1H NMR (300 MHz, CDCl3): d=
7.11–7.37 (m, 34H; ArH), 6.97–7.11 (m, 4H; ArH), 6.81 (s, 4H; core-
ArH), 6.21–6.43 (m, 18H; ArH), 3.72–4.05 (m, 28H; ArOCH2), 2.59–2.93
(m, 28H; ArCH2+PhCH2), 2.12–2.33 (m, 12H; PCH2), 1.94–2.12 (m,
28H; ArCH2CH2+PhCH2CH2), 1.08–1.34 ppm (m, 18H; PCH2CH3);
13C NMR (75 MHz, CDCl3): d=160.2, 153.2, 144.0, 143.9, 143.8, 141.5,
General procedure for the synthesis of poly(dendrimer)s Pt-S-Gn and
Pt-L-Gn: A mixture of the S-Gn or L-Gn dendrimers (1.0 equiv, 8.7 mm),
trans-[PtACHTUNGTRENNUNG(PEt3)2Cl2] (1.0 equiv) in CHCl3/iPr2NH (v/v 1:1) was frozen in a
sealed tube by liquid N2 and degassed with N2 (3 times). The mixture was
allowed to warm to 258C followed by the addition of CuI (0.5 equiv).
The mixture was then heated at 408C for 2 days. The solvent was concen-
trated in vacuo, the residue was dissolved in a minimum amount of
CHCl3 and filtered through a short pad of alumina, and the target poly-
ꢀ
138.3, 130.9, 128.5, 128.4, 128.2, 126.5, 125.9, 115.4, 109.2 (C CPt), 107.4,
2
ꢀ
107.1, 106.8 (t, J
32.7, 32.5, 32.2, 32.0, 31.4, 30.9, 30.7, 16.3 (quintet like, J
G
1
(C,P)=17.5 Hz;
PCH2), 8.4 ppm; 31P NMR (122 MHz, CDCl3): d=(major peak): 11.1 (1J-
(Pt,P)=2380 Hz); (minor peaks): 15.0 (1J(Pt,P)=2390 Hz), 8.7 ppm (1J-
(Pt,P)=2330 Hz).
Pt-L-G3: Starting from L-G3 (194 mg, 0.047 mmol), trans-[Pt
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
AHCTUNGRTEG(NNUN PEt3)2Cl2]
AHCTUNGTRENNUNG
(65.8 mg, 0.131 mmol), and CuI (12.5 mg, 0.066 mmol) in CHCl3/iPr2NH
(v/v 1:1, 15 mL), the polymer (148 mg, 95%; 107 mg, 69% after precipita-
tion) was obtained as a yellow powder. 1H NMR (300 MHz, CDCl3): d=
7.15–7.55 (m, 18H; ArH), 6.89 (s, 4H; core-ArH), 6.67 (s, 4H; ArH),
6.55 (s, 2H; ArH), 5.03 (s, 4H; ArCH2O), 4.97 (s, 4H; ArCH2O), 4.95 (s,
4H; ArCH2O), 1.97–2.31 (m, 12H; PCH2), 1.03–1.37 ppm (m, 18H;
PCH2CH3); 13C NMR (75 MHz, CDCl3): d=160.23, 160.16, 153.1, 139.8,
ACHTUNGTRENNUNG(PEt3)2Cl2]
(23.5 mg, 0.047 mmol), and CuI (4.5 mg, 0.023 mmol) in CHCl3/iPr2NH
(v/v 1:1, 5.4 mL), the product (204 mg, 95%; 193 mg, 90% after precipi-
tation) was obtained as a yellow powder. 1H NMR (300 MHz, CDCl3):
d=6.97–7.36 (m, 78H; ArH), 6.82 (s, 4H; core-ArH), 6.17–6.46 (m, 42H;
ArH), 3.73–4.03 (m, 60H; ArOCH2), 2.57–2.92 (m, 60H;
ArCH2+PhCH2), 1.88–2.33 (m, 72H; PCH2+ArCH2CH2+PhCH2CH2),
ꢀ
136.8, 133.3, 131.1, 128.6, 128.1, 127.6, 127.5, 115.8, 109.3 (C CPt), 108.4
13
ꢀ
(t, 2J
(C,P)=14.3 Hz; C CPt), 106.4, 106.3, 101.5, 70.6, 70.1, 16.4 (quintet
E
1.06–1.40 ppm (m, 18H; PCH2CH3); C NMR (75 MHz, CDCl3, the C
ꢀ
1
CPt signal was too weak to be observed): d=160.3, 160.2, 153.2, 144.0,
like, J
(major peak): 11.1 (1J
2390 Hz), 8.6 ppm (1J
(Pt,P)=2320 Hz).
Pt-S-G2: Starting from S-G2 (211 mg, 0.131 mmol), trans-[Pt
(C,P)=17.5 Hz; PCH2), 8.4 ppm; 31P NMR (122 MHz, CDCl3): d=
143.94, 143.88, 141.6, 138.4, 130.9, 128.6, 128.5, 128.2, 126.0, 115.4, 109.2
ACHTUNGTRENNUNG ACHTUNGTRENNUNG
(Pt,P)=2370 Hz); (minor peaks): 14.9 (1J
ꢀ
(C CPt), 107.2, 99.0, 67.6, 66.93, 66.86, 32.5, 32.2, 32.0, 30.9, 30.8, 30.7,
ACHTUNGTRENNUNG
16.4 (quintet like, 1J
(C,P)=17.7 Hz; PCH2), 14.8, 14.6, 14.3, 8.45, 8.4,
T
8.1 ppm; 31P NMR (122 MHz, CDCl3): d=(major peak): 11.0 (1J
ACHTUNGTRENNUNG
(65.8 mg, 0.131 mmol), and CuI (12.5 mg, 0.066 mmol) in CHCl3/iPr2NH
(v/v 1:1, 15 mL), the product (236 mg, 88%; 218 mg, 81% after precipita-
tion) was obtained as a yellow powder. 1H NMR (300 MHz, CDCl3): d=
7.17–7.51 (m, 38H; ArH), 6.87 (s, 4H; core-ArH), 6.67 (s, 12H; ArH);
6.49–6.60 (m, 6H; ArH), 5.02 (s, 12H; ArCH2O), 4.96 (s, 12H;
ArCH2O), 4.93 (s, 4H; ArCH2O), 1.97–2.31 (m, 12H; PCH2), 1.08–
1.32 ppm (m, 18H; PCH2CH3); 13C NMR (75 MHz, CDCl3): d=160.2,
160.1, 153.1, 139.8, 139.3, 139.2, 136.8, 133.3, 131.0, 128.6, 128.0, 127.6,
2380 Hz); (minor peaks): 14.9 (1J(C,P)=2390 Hz), 8.6 ppm (1J
ACTHNUGTRENNUGN ACHTUNGTRENNUNG
2330 Hz).
Acknowledgement
127.5, 115.8, 109.3 (C CPt), 108.4 (t, 2J
G
ꢀ
ꢀ
We thank the Research Grants Council, Hong Kong SAR, for financial
support (Project No: 400404).
101.6, 70.5, 70.1, 16.4 (quintet like, 1J
G
31P NMR: (122 MHz, CDCl3): d=(major peak): 11.2 (1J
2370 Hz); (minor peak): 8.7 (1J
(Pt,P)=2320 Hz).
Pt-S-G3: Starting from S-G3 (155 mg, 0.047 mmol), trans-[Pt
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
[1] For recent examples, see: a) L. M. Passeno, M. E. Mackay, G. L.
Baker, R. Vestberg, C. J. Hawker, Macromolecules 2006, 39, 740–
746; b) K. T. Kim, M. A. Winnik, I. Manners, Soft Matter 2006, 2,
(23.5 mg, 0.047 mmol), and CuI (4.5 mg, 0.023 mmol) in CHCl3/iPr2NH
(v/v 1:1, 5.4 mL), the product (173 mg, 99%; 162 mg, 93% after precipi-
Chem. Eur. J. 2009, 15, 2278 – 2288
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2287

H.-F. Chow et al.
957–965; c) B.-K. Cho, A. Jain, S. M. Gruner, U. Wiesner, Chem.
Commun. 2005, 2143–2145.
[5] H.-F. Chow, C.-F. Leung, W. Li, K.-W. Wong, L. Xi, Angew. Chem.
2003, 115, 5069–5073; Angew. Chem. Int. Ed. 2003, 42, 4919–4923.
[6] K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 1975, 16,
4467–4470.
[7] C. Hawker, J. M. J. Frꢁchet, J. Am. Chem. Soc. 1990, 112, 7638–7647.
[8] a) A. S. Hay, J. Org. Chem. 1960, 25, 1275–1276; b) A. S. Hay, J. Org.
Chem. 1962, 27, 3320–3321.
[2] For recent reviews, see: a) A. Zhang, L. Shu, Z. Bo, A. D. Schlꢃter,
Macromol. Chem. Phys. 2003, 204, 328–339; b) A. D. Schlꢃter, C. R.
Chim. 2003, 6, 843–851; c) A. D. Schlꢃter, Top. Curr. Chem. 2005,
245, 151–191; d) H. Frauenrath, Prog. Polym. Sci. 2005, 30, 325–384.
[3] R. N. Ganesh, J. Shraberg, P. G. Sheridan, S. Thayumanavan, Tetrahe-
dron Lett. 2002, 43, 7217–7220.
[9] A.-D. Schlꢃter, Top. Curr. Chem. 1998, 197, 165–191.
[4] a) G. R. Newkome, C. N. Moorefield, F. Vçgtle, Dendritic Molecules.
Concepts. Syntheses. Perspectives, Wiley-VCH, Weinheim, 1996, chap.
9; b) G. R. Newkome, C. N. Moorefield, F. Vçgtle, Dendrimers and
Dendrons. Concepts, Syntheses, Applications, Wiley-VCH, Weinheim,
2001, chap. 9.
Received: November 6, 2008
Published online: January 16, 2009
2288
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2278 – 2288