Synthesis of cyclic dendronized polymers via divergent "graft- from" and convergent click "graft-to" routes: Preparation of modular toroidal macromolecules

Article abstract of DOI:10.1021/ja2024355

Cyclic dendronized polymers represent a new class of polymers which exhibit a donut-like "toroidal" shape. Two previous unreported methods for preparing this architecture, the divergent "graft-from" and the convergent click "graft-to" approaches, are explored and the resulting products fully characterized. This route is particularly attractive because it enables production of exact linear and cyclic dendronized analogues, enabling direct comparison of their physical properties. In this preliminary work, the divergent "graft from" approach appears to lead to materials with broad PDI at high DP, whereas the "graft to" approach yields more well-defined dendronized cyclic polymers at larger DP. On the basis of reports to date, a combination of click cyclization followed by click "grafting to" provides the most versatile route for the synthesis of cyclic dendronized polymers.

Full text of DOI:10.1021/ja2024355

ARTICLE
pubs.acs.org/JACS
Synthesis of Cyclic Dendronized Polymers via Divergent “Graft-from”
and Convergent Click “Graft-to” Routes: Preparation of Modular
Toroidal Macromolecules
Boyd A. Laurent and Scott M. Grayson
Department of Chemistry, Tulane University, New Orleans, Louisiana 70118, United States
S Supporting Information
b
ABSTRACT: Cyclic dendronized polymers representa new class
of polymers which exhibit a donut-like “toroidal” shape. Two
previous unreported methods for preparing this architecture, the
divergent “graft-from” and the convergent click “graft-to” ap-
proaches, are explored and the resulting products fully char-
acterized. This route is particularly attractive because it enables
production of exact linear and cyclic dendronized analogues,
enabling direct comparison of their physical properties. In this
preliminary work, the divergent “graft from” approach appears
to lead to materials with broad PDI at high DP, whereas the “graft to” approach yields more well-defined dendronized cyclic
polymers at larger DP. On the basis of reports to date, a combination of click cyclization followed by click “grafting to” provides the
most versatile route for the synthesis of cyclic dendronized polymers.
’ INTRODUCTION
macromonomers. The “graft from” approach involves stepwise
growth of dendrons outward from the polymer backbone. This
technique typically suffers from the same weaknesses as diver-
gent dendrimer syntheses, where the number of functionalities
per macromolecule increases exponentially, increasing the prob-
ability of incomplete reactions, as well as undesirable side
reactions, such as cross-linking. However, for highly optimized
dendritic syntheses, such as the polyester dendrons based on bis-
MPA,21ꢀ27 or the polyamides,28 this approach provides a rela-
tively efficient route to well-defined dendronized polymers.2,12ꢀ14
Finally, the “graft to” approach involves a polymer with reactive
functional groups along the backbone that are coupled to pre-
formed dendrons with a complementary reactive focal point.
This method rarely yields complete dendronization for high
generation dendrons because reactive functionalities on the
linear polymer can be sterically blocked by dendrons on
adjacent repeat units partway through the coupling reaction,
in addition to the steric hindrance around the reactive
functionality at the focal point of the dendron. However, for
highly efficient coupling reactions, such as the Cu catalyzed
alkyneꢀazide click cycloaddition (CuAAC) reaction,29 suffi-
cient coverage can be achieved to access materials that exhibit
the unique properties of dendronized polymers.
Dendronized polymers represent a unique class of hyperfunc-
tional polymers which exhibit tunable rigidity and shape in
addition to control over the length of the polymer backbone.1,2
They consist of a polymeric backbone in which each repeat unit
bears a pendant dendron. The size, steric compactness, and
chemical functionality of each dendron on the polymer backbone
play a large role in defining the macromolecule’s shape and
rigidity and therefore, ultimately, its overall physical properties.
For large dendrons, the strong steric repulsion between adjacent
dendrons inhibits conformational flexibility resulting in a rod-like
shape. Dendronized polymers also display a highly functional
periphery as a result of the multiplicity of dendron end-groups
which largely reside at the surface of these cylindrical macro-
molecules. These surface functionalities are amenable to chemical
modification and therefore provide a rapid means to tune
solubility and the strength of other intermolecular interactions.
While linear dendronized polymers have been the subject of
numerous studies, the properties and applications of other
dendronized polymer architectures, such as cyclic dendronized
polymers, remain largely unexplored.
There are three general approaches for the synthesis of
dendronized polymers: the macromonomer route,3ꢀ11 the “graft
from” route,3,12ꢀ16 and the “graft to” route.17ꢀ20 Briefly, the macro-
monomer approach involves polymerization of a monomer that is
already bearing the desired dendritic side chain. This method often
suffers from low degrees of polymerization and extremely poly-
disperse polymers as a result of steric hindrance which discourage
interaction between the propagating polymer active site and
the polymerizable functionality of the unreacted dendronized
Cyclic dendronized polymers are a relatively new poly-
mer architecture which has only been reported within the last
5 years.30 Preliminary reports from our laboratories have demon-
strated for the first time the viability of the divergent route,31
Received: March 31, 2011
Published: July 20, 2011
r
2011 American Chemical Society
13421
dx.doi.org/10.1021/ja2024355 J. Am. Chem. Soc. 2011, 133, 13421–13429
|

Journal of the American Chemical Society
Scheme 1. Synthesis of Cyclic Poly(4-acetoxystyrene) as a Scaffold for Further Functionalization of the Cyclic Architecture
ARTICLE
while Boydston et al. have recently reported the first and
only reported synthesis of cyclic dendronized polymers utilizing
the macromonomer approach.4 Through the use of a specialized
“cyclic” Ru Grubbs type ring-opening metathesis polymerization
method,32ꢀ35 Boydston et al. were able to successfully synthesize
a second generation cyclic dendronized polymer of extremely
high molecular weight, up to 5 MDa, and were also able to
visualize the individual polymer rings via Atomic Force Micro-
scopy (AFM). Although this route is notable in that it provides
access to cyclic dendronized polymers of sufficient degrees of
polymerization (DP) to enable microscopic visualization, broad
utilization of this approach is limited due to the relatively
challenging synthesis of the unique cyclic Ru catalyst and the
products exhibited the relatively broad molecular weight dis-
tributions associated with the ring-opening metathesis route.
Explorations within our group have focused on the divergent
“graft from” and the convergent “graft to” approaches, in part,
because they should offer more synthetic accessibility and
versatility, in that a wide range of preformed cyclic polymers
can be prepared, and then a variety of different dendrons could
be added in subsequent synthetic steps. Recently, a number of
research groups have reported improved methods for the
synthesis of highly pure simple cyclic polymer backbones
consisting of styrenics, acrylates, methacrylates, acrylamides,
butadienes, and a variety of polyesters including poly-
(caprolactone) and poly(lactide).36ꢀ38 Within our laboratory,
a technique has been developed for the generation of high
purity cyclic polymers employing the slow addition technique
of R-alkynyl,ω-azido-heterodifunctionalized linear poly-
mer utilizing highly efficient (CuAAC) coupling.39ꢀ43 This
approach offers excellent versatility in both tolerance to func-
tional groups present on the polymer as well as access to a broad
range of molecular weight, yet maintains the narrow polydisper-
sity of the linear polymerization techniques used (e.g., atom
transfer radical polymerization (ATRP), radical additionꢀfrag-
mentation chain transfer (RAFT) polymerization, and ring-open-
ing polymerization (ROP)) which have already been the subjectof
extensive optimization. In this technique, the polymer solution is
added at a slow, controlled rate via syringe pump to a solution
of Cu(I)Br “click” catalyst to maintain an infinitesimally small
concentration of linear precursor throughout the course of the
cyclization. Due to the extremely high efficiency of the “click”
reaction between a terminal alkyne and an azide,29 the cycliza-
tion reaction affords near quantitative yields of the cyclic
polymers, removing the need for tedious purification techni-
ques. As a result of the functional group compatibility of this
approach, the facile incorporation of side chain functionality
enables the attachment of dendrons via either the “graft from”
or the “graft to” approaches.
Herein, we report the first detailed synthesis of cyclic den-
dronized polymers utilizing both the “graft from” and “graft to”
approaches. Each approach produces a styrene-based cyclic
polymer backbone with dendritic polyester side chains. In
the divergent “graft from” approach, the dendron is grown in a
stepwise fashion utilizing iterative esterification and deprotection
chemistry. The alternative “graft to” approach consists of the
“clicking” of a library of different generation dendrons bearing a
single azido-group at the focal point onto a polymer backbone
with alkyne side-chains on each repeat unit. Thismethodology was
performed on both linear and cyclic poly(pent-4-ynoyloxy)styrene
(c-PHS-yne) in order to provide a direct comparison of exact
linear and cyclic analogues.
’ RESULTS AND DISCUSSIONS
Divergent “Graft-from” Dendronization of Cyclic Macro-
cycles. The synthesis of cyclic polymer cores for dendronization
utilized the “click” cyclization technique reported recently in our
laboratories.39 4-Acetoxystyrene was chosen as the monomer
unit because poly(4-acetoxystyrene) (PAS) can be readily hydro-
lyzed to poly(4-hydroxystyrene) (PHS), which is an ideal sub-
strate for divergent dendronization12 and it has been shown to
polymerize via Atom Transfer Radical Polymerization (ATRP)
in a controlled fashion to produce polymers with low molecular
weight dispersities.44
4-Acetoxystyrene was polymerized via ATRP using Cu(I)Br
as a catalyst initiating from propargyl 2-bromoisobutyrate 1, a
tertiary R-bromo ester initiator bearing an alkyne to produce a
well-defined polymer, 2, bearing a bromide end-group. The
bromo functionality was then converted to an azido end-group
13422
dx.doi.org/10.1021/ja2024355 |J. Am. Chem. Soc. 2011, 133, 13421–13429

Journal of the American Chemical Society
ARTICLE
by reaction with sodium azide in DMF and the end-modified
polymer 3 was isolated by precipitation into hexanes. The linear
polymer precursors were then cyclized via a highly efficient
intramolecular “click” reaction between the alkynyl group on
the R-initiating end of the polymer and the azido group on the
ω-terminus of the polymer via a Cu(I)Br catalyst (Scheme 1) to
yield macrocyclic polymer 4.
By using a syringe pump to regulate and optimize the slow
addition of the linear precursor 3, the reactant could be converted
to the corresponding cyclic analogue 4 utilizing high dilution to
encourage the intramolecular cyclization over intermolecular
coupling. The GPC traces of the starting material and crude
reaction mixture verified that the linear precursors were con-
verted in a nearly quantitative fashion (Figure 1). A shift to a
longer retention time in the GPC of the cyclic polymer was
consistent with a more compact cyclic architecture and con-
firmed that the intramolecular coupling had occurred.45 The
resultant c-PAS 4 was then deacetylated either via an acid
catalyzed hydrolysis of the phenolic esters using a weakly acidic
solution of 1ꢀ2% v/v H2SO4 in 4/3 THF/MeOH
(∼0.2ꢀ0.4M) or a weakly basic KOH in H2O/MeOH solution
to form the c-PHS precursor, 5, that is necessary for the divergent
dendronization procedure.12,15 This reaction was monitored via
1H NMR by confirming the complete disappearance of the
acetate proton signal at 2.1 ppm. Additionally, a control experi-
ment was performed to prove that the linking pivalate lactone
ester in the backbone of the cyclic polymer was unhydrolyzed
and that ring scission had not occurred to regenerate of the linear
precursor (Figure S1 of the Supporting Information, SI). The
retention of the cyclic architecture was proven by subsequent
reacetylation of the PHS pendant alcohols on the backbone
and overlaying the resultant GPC trace onto the original cyclic
polymer. The resulting PAS polymer exhibited a monomodal
GPC trace that overlaid exactly onto the c-PAS, with a notably
longer retention time than the l-PAS, confirming the cyclic archi-
tecture was retained. It is believed that the observed selective
hydrolysis of ester side chains results from their phenolic nature
Figure 1. Representative GPC traces exhibiting longer retention time for
cyclic poly(4-acetoxystyrene) 4 (red) and linear poly(4-acetoxystyrene)
3 (blue).
Scheme 2. Divergent Dendronization Process Used for Synthesis of Cyclic Dendronized Polymers Employing an Iterative Ester
Anhydride Coupling and Pd-Catalyzed Hydrogenolysis Deprotection Step
13423
dx.doi.org/10.1021/ja2024355 |J. Am. Chem. Soc. 2011, 133, 13421–13429

Journal of the American Chemical Society
ARTICLE
Table 1. Agreement between Theoretical Molecular Weight
and MALDI-TOF MS as Well as GPC Data for Cyclic
Dendronized Polymers
aromatic resonances (∼7.2ꢀ7.6 ppm) as well as the acetal
hydrogen (5.4ꢀ5.5 ppm) via 1H NMR. Additional iterations of
the nearly quantitative dendron growth and deprotection reac-
tions were verified by a transition to shorter retention times in the
GPC for compounds 7, 9, and 11 (Figure 2), and by a doubling of
Figure 2. GPC traces of differing generation of poly(ester) based cyclic
dendronized polymers produced via an iterative divergent “graft from”
dendronization technique.
1
the H NMR integration values for the benzylidene protecting
groups relative to the protons of the styrenic backbone (6.0ꢀ7.2
ppm) (Figure 3), as well as a doubling of the MWs in the
MALDI-TOF mass spectra (Table 1). GPC traces of each of the
cyclic dendronized polymers also show retention of a relatively
low polydispersity throughout the dendronization procedure
highlighting both the efficiency of this anhydride esterification/
deprotection dendronization process and well-defined nature of
the products. Molecular weights calculated by the GPC elution
volumes were much lower than theoretical molecular weights,
but this is expected due to the highly branched and compact
nature of the dendronized units (Table 1). This coupling
procedure gave excellent results for first, second, and third
generation cyclic dendronized polymers with degrees of polym-
erization of less than ∼30 and upon deprotection, the third
generation hydroxylated polymer with 8 hydroxyl units per
repeat unit, exhibited modest solubility in water. However, while
cyclic polymers remained monomodal at low DP (<30) through
the dendronization process to the third generation, larger DP
polymers (∼38) that maintained low PDIs for the first two
generations, exhibited a loss of monomodality by GPC at higher
generations of dendritic growth as noted by an impurity with
shorter retention time (Figure S2 of the SI). The appearance of a
high molecular weight shoulder has previously been reported
during the divergent dendronization of linear polymer cores at
high molecular weights using the same benzylidene protected
anhydride monomer, and is hypothesized to result from either a
covalent cross-linking side-reaction or a noncovalent aggregated
assembly in these large, highly functional polymers.12 We have
also observed similar results in our synthesis when attempting to
increase the degree of dendronization beyond the third genera-
tion for polymers with DP > 30. Divergent dendronization was
also explored using acetonide protected bis-MPA monomers, but
exhibited a similar loss of monomodality at high molecular
weights. In order to explore an alternative route to making
well-defined cyclic polymers, as well as expedite the synthetic
procedure, an alternative “graft to” approach was investigated
where preformed dendrons are attached to the cyclic core via a
CuAAC coupling reaction.
Figure 3. Selected 1H NMR region of first (bottom), second (middle),
and third (top) generation cyclic dendronized polymers produce via
divergent dendronization showing near quantitative functionalization of
the polymeric backbone. Highlighted are the resonances corresponding
to the aromatic protons of the c-PHS core (6.0ꢀ7.15 ppm) and the
acetal proton of the benzylidene acetal protecting group (5.35ꢀ5.6
ppm).
as well as the substantial steric hindrance around the pivalate
linker on the backbone.
Divergent dendronization of the resultant c-PHS 5 was carried
out via an iterative procedure of esterification and hydrogenolysis
(Scheme 2). First, the alcohol on each repeat unit of 5 was
esterified using a benzylidene protected acid anhydride monomer
6 to produce the first generation cyclic dendronized polymer 7.
The cyclic benzylidene acetal was then deprotected via Pd-
catalyzed hydrogenolysis to give the corresponding first genera-
tion cyclic dendronized polymer 8 with two alcohol surface
functionalities per repeat unit. Benzylidene deprotection of the
dendron corona was monitored by complete loss of benzylidene
Convergent “Graft to” Dendronization of Cyclic Macro-
cycles. The “graft to” approach utilized the same synthetic route
to prepare the c-PHS 5 core as the “graft from” approach but
carries out the iterative dendritic synthesis on the free dendrons
before they are coupled to the core. It has been demonstrated
13424
dx.doi.org/10.1021/ja2024355 |J. Am. Chem. Soc. 2011, 133, 13421–13429

Journal of the American Chemical Society
ARTICLE
Scheme 3. Synthetic Methods Outlining the Process for an Exactly Comparable Linear and Cyclic Library of Polymers as Well as
Functionalization of the Poly(4-hydroxystyrene) Backbone with 4-Pentynoic Acid for Click Grafting
Figure 5. Linear 14 (blue) and cyclic 12 (red) PHS-alkyne scaffolds
used for convergent click dendronization.
Figure 4. 1H NMR and integration values of the cyclic alkyne func-
tionalized poly(4-hydroxystyrene), c-PHS-yne (12) indicating near
quantitative functionalization of the PHS backbone.
whereas the linear PHS core, 13, was produced through reduc-
tion of the acetate functionalities along the polymer backbone as
well as the ester functionality of the initiator with LiAlH4. These
strong reducing conditions were employed instead of hydrolytic
conditions because of the uncertainty of potential side-reactions
which may result from the R-carboxylic acid functionality in
subsequent steps that would be produced during a simple acidic
hydrolysis of the initiating ester.
previously that linear dendronized polymers can be produced
with high grafting efficiency via “click” coupling of azido-
functionalized dendrons as large as the third generation onto
alkyne functionalized linear polymer backbones in the presence
of a Cu(I) catalyst.17 Additionally, cyclic polymer brushes have
also been reported through a click “graft to” method.46 This
alternative approach, much like the previous “graft from” ap-
proach, provides rapid synthetic access to a family of cyclic
dendronized polymers, as well as an exactly analogous family of
linear dendronized polymers (if the dendronization procedure is
carried out on the same batch of linear core 2 before the
cyclization step) (Scheme 3). These are attractive synthetic
routes for elucidating the effect of the cyclic architecture on
physical properties, as the linear and cyclic analogs should have
nearly identical molecular weights and side chain structure,
differing only in molecular topology.
In order to adapt the parallel linear and cyclic “graft to”
approach to our cyclization methodology, l-PHS 13 and c-PHS
5 were each functionalized with an alkynyl moiety through
reaction of the pendant alcohols of PHS with 4-pentynoic
acid using N-(3-dimethylaminopropyl)-N0-ethylcarbodiimide
hydrochloride (EDC) coupling agent (Scheme 3). A com-
1
bination of MALDI-TOF MS and H NMR integration values
comparing the aromatic protons of the PHS core (6.20ꢀ7.00 ppm)
and the methylene units of the formed alkynyl esters (2.70ꢀ2.90
ppm) of both the cyclic, 12, and linear, 14, alkyne functionalized
polymers indicated near quantitative functionalization of the polymer
backbone (Figure 4). Additionally, GPC traces verified that both
linear and cyclic alkynylated polymer products maintained a narrow
The cyclic PHS core, 5, was produced using an identical
cyclization and deprotection procedure as described above,
13425
dx.doi.org/10.1021/ja2024355 |J. Am. Chem. Soc. 2011, 133, 13421–13429

Journal of the American Chemical Society
ARTICLE
Scheme 4. Divergent Dendronization of 3-Azido-propan-1-ol (16) To Produce a Library of Azido Functionalized Dendrons
(18aꢀd)
polydispersity and a shorter elution volume compared to their acetate
protected counterparts 4 and 2, respectively. Furthermore, GPC was
also able to still resolve differences between the cyclic alkynylated
polymer 12 and the less compact linear analogue 14, as noted by the
shorter elution volume of the linear analog (Figure 5).
example, in the case of the third generation 18c, only two peaks
were observed in the MALDI-TOF spectrum, the first corre-
sponding to the expect mass of the sodium adduct, 1096.47
([M + Na]+ theoretical m/z = 1096.50 Da), and the second
corresponded to an in-source metastable fragmentation (loss
of ꢀN2) that commonly occurs for polymeric azides during
MALDI-TOF MS characterization.48 Lower molecular weight
impurities that might result from degradation or incomplete
couplings were not observed which indicate that all coupling
and deprotection reactions were carried out to completion. In
addition, IR spectra of all of the azido-functionalized dendrons
A library of dendrons with a single azide functionality at the
focal point were then prepared via a divergent dendronization
procedure using an acetonide protected bis-MPA acid anhydride
monomer (Scheme 4).21,47 The azido group was first introduced
by the reaction of NaN3 with commercially available 3-bromo-
propanol 15 to produce 3-azido-propanol 16 which was then
esterified with acetonide protected bis-MPA acid anhydride, 17,
to produce the protected first generation azido functionalized
dendron 18a. The acetonide groups on this dendron could then
be quantitatively removed using a solid phase acid exchange resin
in MeOH to produce the first generation deprotected azido
functionalized diol dendron 19a. The completion of the depro-
tection reaction could be easily verified by the loss of acetonide
resonances at 1.36 and 1.41 ppm in the 1H NMR spectrum. This
procedure was repeated in an iterative fashion in order to
produce the second (18b), third (18c), and fourth (18d)
generation azido functionalized dendrons. The monodisperse
nature of the MALDI-TOF mass spectra of the resultant third
18c and fourth 18d generation dendron confirmed that complete
coupling and deprotection had occurred at each step. For
exhibited a clear azido stretching frequency at 2100 cmꢀ1
,
confirming retention of the focal azide functionality throughout
the dendronization process.
These azido-functionalized dendrons 18aꢀd were then
“clicked” to the alkynyl functionalized linear (14) and cyclic
(12) polymer cores (Scheme 5). This “click” grafting approach
was investigated for first (18a), second (18b), third (18c), and
fourth (18d) generation azido functionalized dendrons, using a
1.1/1 ratio of azide/alkyne with a Cu(I)Br/PMDETA catalyst
in DMF. For the cyclic core, the “click” grafted products 20aꢀc
exhibited increasingly shorter GPC retention times with respect
to the increasing generation of dendritic side chains (Figure 6).
However, reaction with the fourth generation azido-functionalized
dendron exhibited only minimal coupling onto the cyclic polymer
13426
dx.doi.org/10.1021/ja2024355 |J. Am. Chem. Soc. 2011, 133, 13421–13429

Journal of the American Chemical Society
ARTICLE
Scheme 5. Convergent “Graft to” Click Coupling Route
of a Library of Azido-Functionalized Dendrons (18aꢀc)
onto Both Linear (14) and Cyclic (12) Alkynylated Polymer
Cores
core, as judged by GPC, even when using microwave irradiation,
elevated temperature, and extended reaction times. This result
coincides with the first reported use of CuAAC to dendronize
linear polymers, in which Frꢀechet-type aryl ether dendrons
exhibited poor coupling efficiencies above the third generation17
suggesting that the azido-functionality is sufficiently buried within
the dendritic structure at the fourth generation so as to prevent
reaction with the polymeric alkynes.
However, unlike the divergent “graft from” approach, the GPC
traces remain monomodal even for the higher molecular weight
cores, and with dendrons as large as the third generation,
suggesting this click “graft to” approach can overcome the
complications observed for the divergent “graft from” approach
for higher molecular weight dendronized polymers. Similar
results were also noted for the click “graft to” approach for the
analogous linear dendronized polymer products (Figure S3 of the
SI).
A significant advantage to preparing exactly analogous linear
and cyclic dendronized polymers is the ability to verify and explore
changes in physical properties relative to the architecture of the core
that has been dendronized. For example, at the first generation, the
cyclic [G1]-dendronized polymer 20a still shows a notably longer
retention time than its linear counterpart 21a, resulting from the
relatively confined configuration of the dendronized cyclic polymer as
a result of its ring-like topology (Figure 7). A similar trend is also
observed, but less pronounced for the second generation of grafted
polymers, namely cyclic [G2]-dendronized polymer 20b and its linear
analogue 21b. However, the difference is not discernible for the third
generation presumably because size of the dendritic side chains for
this generation begins to approach the length of the polymer back-
bones, (DP = 38), yielding similar globular structures regardless of
backbone architecture.
These polymers were also characterized in detail by 1H NMR in
1
order to quantify the extent of grafting. H NMR spectra of each
generation of the cyclic dendronized polymers (20aꢀc) are shown
in Figure 8. The coupling efficiency was calculated by comparing the
integrated values of the protons of the polymeric backbone (aromatic
protons at 6ꢀ7 ppm) and the proton found on the formed triazole
ring (∼7.5 ppm). The values for the calculated coupling efficiency
are shown in Table 2. It should also be noted that Cu(0) was used as
an additive to the “click” reaction in an effort to increase the
concentration of the active Cu(I) species for the coupling of both
second and third generation dendritic azides onto the polymer
backbone. The addition of Cu(0) to the reaction resulted in much
higher coupling efficiencies, typically around 10ꢀ20% higher than
without Cu(0), which have been attributed to the ability of Cu(0) to
Figure 6. GPC traces of cyclic “click” dendronized polymers 20aꢀc:
c-PHS-alkyne 12 (red), c-PHS-triaz-[G1]-An 20a (green), c-PHS-
triaz-[G2]-(An)2 20b (purple), and c-PHS-triaz-[G3]-(An)4 20c
(orange).
Figure 7. GPC traces of linear (blue) and cyclic (red) first (a) and second (b) generation “click” dendronized polymers, respectively.
13427
dx.doi.org/10.1021/ja2024355 |J. Am. Chem. Soc. 2011, 133, 13421–13429

Journal of the American Chemical Society
ARTICLE
Table 3. Theoretical and Observed MALDI-TOF MS Data
for “Graft to” Coupling Process at Each Generation for Both
the Linear (21aꢀc) and Cyclic (20 aꢀc) Dendronized
Polymers
Figure 8. 1H NMR spectra of cyclic “graft to” dendronized polymers
of first generation 20a (top), second generation 20b (middle), and
third generation 20c (bottom) highlighting growth of dendrons with
respect to polymer backbone with increasing generation. Also noted is
the presence of the aromatic triazole proton which was used to
determine the coupling efficiency of the “graft to” process present at
7.5 ppm.
the third generation dendronization of the cyclic backbone yielding
72% coupling without Cu(0) and 88% with Cu(0).
MALDI-TOF mass spectra also verified the high extent of
CuAAC coupling through the third generation of dendrons, based on
molecular weight calculations (Figure 9). Because MALDI-TOF MS
enables measurement of the absolute molecular weight, the number
average molecular weights calculated from the mass spectra nearly
double for the first (20a) to second generation (20b) and double
again at the third generation (20c). The Mn for MALDI-TOF values
observed match closely to the predicted values (Table 3) and in the
case of the third generation, the spacing of the crudely resolved peaks
agrees with the predicted repeat unit of 1275 Da.
Table 2. GPC Data and Coupling Efficiency at Each
Generation for Both the Linear (21 aꢀc) and Cyclic (20 aꢀc)
Dendronized Polymers
’ CONCLUSIONS
We have developed synthetic methods for the preparation of
cyclic dendronized polymers through two routes, namely a
divergent “graft from” approach and a convergent “graft to”
approach. Both methods rely on the use of a cyclic poly-
(4-acetoxystyrene) produced via a CuAAC end-to-end intramo-
lecular cyclization reaction performed at high dilution and
therefore provide access to exactly comparable linear and cyclic
analogues. The “graft from” route was successfully performed via
an iterative anhydride esterification reaction followed by catalytic
hydrogenolysis of the benzylidene protecting groups on the
surface of the dendronized polymers. This method allows for
the preparation of cyclic dendronized materials up to the
third generation for polymer exhibiting a DP of less than 30;
however, for larger polymers, high generation materials with
narrow polydispersity were difficult to obtain. The convergent
“graft to” method was also explored to produce cyclic polymers
by first functionalizing the cyclic polymer backbone “core” with
an alkynyl group and subsequently coupling to a library of azido-
functionalized dendrons. This method allows for the efficient
preparation of up to third generation cyclic dendronized materi-
als showing approximately 90% coupling efficiency at the third
generation without loss of monomodality at higher molecular
weights. Additionally, a set of analogous dendronized linear
polymers can be prepared, and comparison of the GPC retention
times exhibited the expected shift to more compact configura-
tions for the cyclic polymers when compared to their linear
analogues through the second generation. These results suggest
the “graft-to” route is the most versatile of the three routes,
in enabling control of the macrocycle diameter, degree of
dendronization, and overall molecular weight, while yielding
extremely narrow polydispersity cyclic dendronized polymers.
* Cu(0) was added during CuAAC.
Figure 9. MALDI-TOF mass spectra of first [G1] 20a (top), second
[G2] 20b (middle), and third 20c [G3] (bottom) generation cyclic
“graft to” dendronized polymers exhibiting an increase in molecular
weight upon coupling of higher generation azido-functionalized den-
drons (18aꢀc). MALDI-TOF mass spectra of analogous series of linear
“graft to” dendronized polymers (21aꢀc) can be found in the Support-
ing Information (Figure S4).
regenerate active Cu(I) from the oxidized Cu catalyst. For example,
in the case of the second generation dendronization of the linear
backbone, a coupling efficiency of 70% was noted without Cu(0),
whereas the incorporation of Cu(0) during the “click” reaction
showed a coupling efficiency of 90%. Similar results were noted for
13428
dx.doi.org/10.1021/ja2024355 |J. Am. Chem. Soc. 2011, 133, 13421–13429

Journal of the American Chemical Society
ARTICLE
’ ASSOCIATED CONTENT
(21) Wu, P.; Malkoch, M.; Hunt, J. N.; Vestberg, R.; Kaltgrad, E.;
Finn, M. G.; Fokin, V. V.; Sharpless, K. B.; Hawker, C. J. Chem. Commun.
2005, 5775–5777.
(22) Ihre, H.; Padilla De Jesꢀus, O. L.; Frꢀechet, J. M. J. J. Am. Chem.
Soc. 2001, 123, 5908–5917.
(23) Malkoch, M.; Malmstr€om, E.; Hult, A. Macromolecules 2002,
35, 8307–8314.
(24) Gillies, E. R.; Frꢀechet, J. M. J. J. Am. Chem. Soc. 2002,
124, 14137–14146.
(25) Grayson, S. M.; Frꢀechet, J. M. J. Org. Lett. 2002, 4, 3171–3174.
(26) Parrott, M. C.; Marchington, E. B.; Valliant, J. F.; Adronov, A.
J. Am. Chem. Soc. 2005, 127, 12081–12089.
S
Supporting Information. Characterization of all synthetic
b
compounds, data for reacylation of compound 5 as control
experiment, example of GPC traces for high DP cyclic dendro-
nized polymers prepared by the “graft from” route, and MALDI
and GPC data for linear dendronized compounds 21aꢀc. This
material is available free of charge via the Internet at http://
pubs.acs.org.
’ AUTHOR INFORMATION
(27) Yim, S.-H.; Huh, J.; Ahn, C.-H.; Park, T. G. Macromolecules
2007, 40, 205–210.
(28) Tomalia, D. A.; Baker, H.; Dewald, J.; Hall, M.; Kallos, G.;
Martin, S.; Roeck, J.; Ryder, J.; Smith, P. Macromolecules 1986,
19, 2466–2468.
Corresponding Author
sgrayson@tulane.edu
(29) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed.
2001, 40, 2004–2021.
’ ACKNOWLEDGMENT
The authors acknowledge Tulane University, the NSF
(CAREER DMR#0844662) for financial support, the Louisiana
Board of Regents for a Graduate Fellowship (B.A.L), the NSF
(MRI #0619770) for enabling MALDI-TOF and ESI mass
spectral characterization, and ESI-MS assistance from Amelia
Neuberger (Tulane University Department of Chemistry).
(30) Grayson, S. M. Nat. Chem. 2009, 1, 178–179.
(31) Laurent, B. A.; Grayson, S. M. Polym. Prepr. (Am. Chem. Soc.,
Div. Polym. Chem.) 2008, 49, 287–288.
(32) Bielawski, C. W.; Benitez, D.; Grubbs, R. H. Science 2002,
297, 2041–2044.
(33) Bielawski, C. W.; Benitez, D.; Grubbs, R. H. J. Am. Chem. Soc.
2003, 125, 8424–8425.
(34) Boydston, A. J.; Xia, Y.; Kornfield, J. A.; Gorodetskaya, I. A.;
Grubbs, R. H. J. Am. Chem. Soc. 2008, 130, 12775–12782.
(35) Zhang, K.; Lackey, M. A.; Cui, J.; Tew, G. N. J. Am. Chem. Soc.
2011, 133, 4140–4148.
(36) Laurent, B. A.; Grayson, S. M. Chem. Soc. Rev. 2009,
38, 2202–2213.
’ REFERENCES
(1) Schl€uter, A. D.; Rabe, J. P. Angew. Chem., Int. Ed. 2000, 39,
864–883.
(2) Shu, L.; Schl€uter, A. D.; Ecker, C.; Severin, N.; Rabe, J. P. Angew.
Chem., Int. Ed. 2001, 40, 4666–4669.
(37) Hoskins, J. N.; Grayson, S. M. Polym. Chem. 2011, 2, 289–299.
(38) Kricheldorf, H. R. J. Polym. Sci. A, Polym. Chem. 2010,
48, 251–284.
(39) Laurent, B. A.; Grayson, S. M. J. Am. Chem. Soc. 2006, 128,
4238–4239.
(40) Eugene, D. M.; Grayson, S. M. Macromolecules 2008, 41,
5082–5084.
(41) Hoskins, J. N.; Grayson, S. M. Macromolecules 2009, 42,
6406–6413.
(3) Malkoch, M.; Carlmark, A.; Woldegiorgis, A.; Hult, A.; Malmstr€om,
E. E. Macromolecules 2004, 37, 322–329.
(4) Boydston, A. J.; Holcombe, T. W.; Unruh, D. A.; Frꢀechet, J. M. J.;
Grubbs, R. H. J. Am. Chem. Soc. 2009, 131, 5388–5389.
(5) Nystr€om, A.; Malkoch, M.; Furꢀo, I.; Nystr€om, D.; Unal, K.;
Antoni, P.; Vamvounis, G.; Hawker, C.; Wooley, K.; Malmstr€om, E.;
Hult, A. Macromolecules 2006, 39, 7241–7249.
(6) Hawker, C. J.; Frꢀechet, J. M. J. Polymer 1992, 33, 1507–1511.
(7) Malenfant, P. R. L.; Frꢀechet, J. M. J. Macromolecules 2000,
33, 3634–3640.
(42) Qiu, X.-P.; Tanaka, F.; Winnik, F. M. Macromolecules 2007,
40, 7069–7071.
(8) Stocker, W.; Karakaya, B.; Sch€urmann, B. L.; Rabe, J. P.; Schl€uter,
(43) Misaka, H.; Kakuchi, R.; Zhang, C.; Sakai, R.; Satoh, T.;
Kakuchi, T. Macromolecules 2009, 42, 5091–5096.
(44) Gao, B.; Chen, X.; Ivꢀan, B.; Kops, J.; Batsberg, W. Macromol.
Rapid Commun. 1997, 18, 1095–1100.
(45) A trace of similar high MW impurities were observed during the
preparation of cyclic polystyrene. Detailed attempts to characterize this
high MW impurity confirmed the presence of cyclic dimer and is the
subject of a manuscript in preparation.
(46) Zhang, K.; Lackey, M. A.; Wu, Y.; Tew, G. N. J. Am. Chem. Soc.
2011, 133, 6906–6909.
(47) Ihre, H.; Hult, A.; Frꢀechet, J. M. J.; Gitsov, I. Macromolecules
1998, 31, 4061–4068.
(48) Li,Y.;Hoskins, J.N.;Sreerama,S. G.;Grayson,S. M.Macromolecules
2010, 43, 6225–6228.
A. D. J. Am. Chem. Soc. 1998, 120, 7691–7695.
(9) Li, W.; Wu, D.; Schl€uter, A. D.; Zhang, A. J. Polym. Sci. A, Polym.
Chem. 2009, 47, 6630–6640.
(10) Li, W.; Zhang, A.; Feldman, K.; Walde, P.; Schl€uter, A. D.
Macromolecules 2008, 41, 3659–3667.
(11) Zhang, A.; Zhang, B.; W€achtersbach, E.; Schmidt, M.; Schl€uter,
A. D. Chem.—Eur. J. 2003, 9, 6083–6092.
(12) Grayson, S. M.; Frꢀechet, J. M. J. Macromolecules 2001, 34,
6542–6544.
(13) Yin, R.; Zhu, Y.; Tomalia, D. A.; Ibuki, H. J. Am. Chem. Soc.
1998, 120, 2678–2679.
(14) Shu, L.; G€oossl, I.; Rabe, J. P.; Dieter Schl€uter, A. Macromol.
Chem. Phys. 2002, 203, 2540–2550.
(15) Yoshida, M.; Fresco, Z. M.; Ohnishi, S.; Frꢀechet, J. M. J.
Macromolecules 2005, 38, 334–344.
(16) Lee, C. C.; Frꢀechet, J. M. J. Macromolecules 2006, 39, 476–481.
(17) Helms, B.; Mynar, J. L.; Hawker, C. J.; Frꢀechet, J. M. J. J. Am.
Chem. Soc. 2004, 126, 15020–15021.
(18) Desai, A.; Atkinson, N.; Rivera, F.; Devonport, W.; Rees, I.;
Branz, S. E.; Hawker, C. J. J. Polym. Sci. A, Polym. Chem. 2000,
38, 1033–1044.
(19) Mynar, J. L.; Choi, T.-L.; Yoshida, M.; Kim, V.; Hawker, C. J.;
Frꢀechet, J. M. J. Chem. Commun. 2005, 5169–5171.
(20) Shu, L.; Sch€afer, A.; Schl€uter, A. D. Macromolecules 2000,
33, 4321–4328.
13429
dx.doi.org/10.1021/ja2024355 |J. Am. Chem. Soc. 2011, 133, 13421–13429