Synthesis of cyclic (Co)polymers by atom transfer radical cross-coupling and ring expansion by nitroxide-mediated polymerization

Article abstract of DOI:10.1021/ma102313q

A novel approach to prepare cyclic polymers via a combination of atom transfer radical polymerization (ATRP) and atom transfer radical cross-coupling (ATRC) is presented. A functional ATRP initiator possessing an alkoxyamine group was synthesized and used

Full text of DOI:10.1021/ma102313q

240 Macromolecules 2011, 44, 240–247
DOI: 10.1021/ma102313q
Synthesis of Cyclic (Co)polymers by Atom Transfer
Radical Cross-Coupling and Ring Expansion by
Nitroxide-Mediated Polymerization
€
Renaud Nicolay and Krzysztof Matyjaszewski*
Center for Macromolecular Engineering, Department of Chemistry, Carnegie Mellon University,
4400 Fifth Avenue, Pittsburgh, Pennsylvania 15213, United States
Received October 8, 2010; Revised Manuscript Received November 25, 2010
ABSTRACT: A novel approach to prepare cyclic polymers via a combination of atom transfer radical
polymerization (ATRP) and atom transfer radical cross-coupling (ATRC) is presented. A functional
ATRP initiator possessing an alkoxyamine group was synthesized and used to prepare well-defined linear
telechelic R-nitroxy,ω-bromo homo- and block copolymers via ATRP and subsequent thermal deprotection
of the R-terminal nitroxide function. Cyclization reactions were achieved by intramolecular ATRC under
high dilution. Ring expansion of a cyclic polystyrene alkoxyamine prepared by ATRC was conducted via
nitroxide-mediated polymerization (NMP) of styrene and yielded a mixture of linear and cyclic poly-
condensates. This result is discussed in detail and explained by the persistent radical effect (PRE), thermal
initiation, and the occurrence of radical crossover reactions during chain extension via NMP.
Introduction
carbon-carbon bonds polymers. Since the macrocycle is directly
obtained from the linear parent by ring closing, and the size of
Cyclic polymers are fascinating materials for physicists and
theoreticians.1,2 The absence of chain-ends imposes a topological
restriction that results in a variety of physical properties that
significantly distinguish macrocycles from their linear counter-
parts. For example, it has been shown that cyclic polymers have
smaller hydrodynamic volumes and are less viscous than their
linear analogues.1,2 In addition, rheological studies of cyclic poly-
mers, which cannot reptate in the conventional sense, due to a
lack of chain-ends, could help improving the current under-
standing of the fundamentals of polymer chain dynamics.3-5
However, the synthesis of well-defined cyclic polymers is still a
challenge in polymer science.1,2 Cyclic polymers can be obtained
in several different ways. Macrocycles can be directly recovered,
as a more or less important fraction, in polymer systems exposed
to ring-chain equilibria, or they can be synthesized using either
ring expansion or ring closure of linear chains by end-to-end
coupling.
Only few reports describe the formation of large rings by direct
insertion of a monomer into a reactive cyclic precursor.2 Although
very attractive, this approach requires use of polymerization
processes in which chain transfer and exchange reactions between
active species are absent. One of the most striking examples of
macrocycles synthesis by monomer insertion is the ring expansion
of a cyclic carbene ruthenium complex by ring-opening metath-
esis polymerization (ROMP).6-8
linking unit is negligible, this approach provides linear and cyclic
structures with same molar mass. Moreover, the use of living
polymerization techniques for the preparation of the linear
precursors allows control of the molecular weight with narrow
MWD. This is of special importance for direct comparison of the
properties of materials consisting of linear or cyclic chains.
Such an approach has been applied to controlled radical
polymerization (CRP).9 A strategy to prepare macrocycles via
the combination of atom transfer radical polymerization
(ATRP)10-13 and click chemistry14-16 was recently reported.17
Linear PSt precursors were prepared by ATRP of St, using
propargyl 2-bromoisobutyrate, as an initiator, followed by nucleo-
philic substitution of the terminal bromine atom by an azido
group.18 The cyclization reaction was conducted under high
dilution conditions, with CuI/L complexes as catalyst. The
functionalization and the cyclization were essentially quantita-
tive, eliminating the need for purification. This approach was
subsequently extended to the synthesis of cyclic poly(N-
isopropylacrylamide)19 (PNIPAM) and cyclic poly(methyl ac-
rylate)-b-polystyrene (PMA-b-PSt) block copolymers.20 How-
ever, in the latter case, the alkyne moiety of the ATRP initiator
had to be protected with a trimethylsilyl group, which added an
additional deprotection step to the synthesis. Recently, reversible
addition-fragmentation chain transfer (RAFT) polymerization21,22
and nitroxide-mediated radical polymerization (NMP)23,24 were also
combined with Huisgen 1,3-dipolar cycloaddition reaction to pre-
pare cyclic PNIPAM25 and PSt.26-28
The simplest and most appropriate methods for the synthesis
of cyclic polymers, with controlled size and narrow molecular
weight distribution (MWD), are based on the end-to-end chain
coupling of R,ω-hetero-difunctional linear chains in highly dilute
reaction conditions.2 This approach offers several significant
advantages over other approaches as it may be used both for
polymers having in-chain reactive functions and for systems
with no labile linkages in their backbone, such as saturated
Click reactions are perfectly suited for the preparation of cyclic
polymers, due to the near-quantitative yields and high selectivity,
even in the presence of other functional groups. However, the
synthesis of cyclic polymers via a combination of CRP and
azide-alkyne click coupling always requires one or two post-
polymerization functionalization steps in order to introduce the
reactive chain-end groups. Therefore, the search for a synthetic
route that would lead to well-defined cyclic polymers in high
*Corresponding author: Tel þ1-412-268-3209; e-mail km3b@
andrew.cmu.edu.
pubs.acs.org/Macromolecules
Published on Web 12/23/2010
r 2010 American Chemical Society

Article
Macromolecules, Vol. 44, No. 2, 2011 241
Scheme 1. Synthesis of Cyclic Polymers via a Combination of ATRP and ATRC
yields, without postpolymerization functionalization, or with a
single, simple, deprotection step before the cyclization, is still
ongoing.
and the solution was then heated to 40 °C for 18 h. The
solution was filtered and concentrated under vacuum. The
crude product was purified by column chromatography
eluting with hexane/ethyl acetate (95/5 gradually increasing
In the present work, we introduce a novel approach to prepare
cyclic polymers via a combination of ATRP and atom transfer
radical cross-coupling (ATRC)29,30 (Scheme 1). Linear telechelic
R-nitroxy-ω-bromo precursors were prepared via ATRP at
relatively low temperature, using an initiator bearing an alkoxy-
amine group, followed by thermal deprotection of the nitroxide
function. Cyclization reactions were subsequently achieved by
intramolecular ATRC, under high dilution conditions. Well-
defined cyclic PMA, PSt, and cyclic PMA-b-PSt block copolymer
were prepared using this approach.
1
to 5/5) to afford 8.15 g (yield = 80.2%) of a white solid. H
NMR (CDCl3) δ: 7.38-7.00 (m, 5H), 4.80 (q, 1H, J=5.54 Hz),
4.05-3.88 (m, 1H), 1.87 (td, 1H, Jd=10.2 Hz, Jt=3.2 Hz), 1.74 (td,
1H, Jd=10.3 Hz, Jt=3.2 Hz), 1.58-1.39 (m, 2H), 1.52 (d, 3H, J=
5.63 Hz), 1.36 (s, 3H), 1.25 (s, 3H), 1.11 (s, 3H), 0.71 (s, 3H).
13C NMR (CDCl3) δ: 145.60, 128.19, 127.08, 126.80, 83.44,
63.47, 60.35, 60.14, 49.07, 48.97, 34.60, 34.28, 23.49, 21.40.
Synthesis of the 2-Bromopropionate Alkoxyamine, 4. 1-Phe-
nylethylhydroxy-TEMPO alkoxyamine (3) (3 g, 10.8 mmol) and
pyridine (1.26 mL, 15.5 mmol) were dissolved in dry dichloro-
methane (17.5 mL). The reaction mixture was cooled in an
ice-water bath, and a solution of 2-bromopropionyl bromide
(1.42 mL, 13.5 mmol) in dry dichloromethane (7.5 mL) was
slowly added while stirring. The mixture was stirred in the
cooling bath for 1 h and then at room temperature for 16 h.
The excess 2-bromo-2-methylpropionyl bromide was neutra-
lized with 0.2 mL of water, and the reaction mixture was then
poured into a solution of hydrochloric acid (100 mL, 0.3 M). The
organic layer was washed with a solution of sodium hydroxide (100
mL, 0.3 M), dried over magnesium sulfate, and concentrated under
vacuum. The product was purified by flash column chromato-
graphy on basic alumina using THF as solvent to afford 4.460 g
(quantitative yield) of yellow oil. 1H NMR (CDCl3) δ: 7.37-7.17
(m, 5H), 5.11-4.96 (m, 1H), 4.78 (q, 1H, J=5.58 Hz), 4.31 (q, 1H,
J = 5.74 Hz), 1.97-1.83 (m, 2H), 1.80 (d, 3H, J = 5.78 Hz),
1.70-1.54 (m, 2H), 1.49 (d, 3H, J=5.48 Hz), 1.35 (s, 3H), 1.27 (s,
3H), 1.13 (s, 3H), 0.68 (s, 3H). 13CNMR(CDCl3) δ: 169.97, 145.41;
128.23, 127.18, 126.84, 83.55, 68.86, 60.08, 44.38, 44.15, 40.58,
34.50, 34.19, 23.37, 21.70, 21.24.
General Procedure for the ATRP of MA and St with the
2-Bromopropionate Alkoxyamine 4 or a PMA Macroinitiator. In
a typical experiment, CuBr (71.7 mg, 0.5 mmol) and CuBr2 (11.2
mg, 0.05 mmol) were charged to a flask. The flask was deoxy-
genated by purging with N2 for 30 min. Deoxygenated acetone
(3.25 mL) and PMDETA (115 μL, 0.55 mmol) were added. After
5 min, deoxygenated methyl acrylate (6.5 mL, 72 mmol) and
2-bromopropionate alkoxyamine, 4 (412 mg, 1 mmol), were
added. An initial sample was taken, and then the flask was
placed in an oil bath thermostated at 40 °C for 20.5 h. The
polymerization was stopped via exposure to air and dilution in
THF (Mn=2620, Mw/Mn=1.09, conversion=38.0%).
General Procedure for the Synthesis of r-Nitroxy-ω-Bromo
PMA, PSt, and PMA-b-PSt. In a typical experiment, TEMPO
(6.25 g, 40 mmol), R-aminooxy-ω-bromo PMA (1.05 g, 0.4
mmol, Mn = 2620, Mw/Mn = 1.09), and anisole (15 mL) were
charged to a 25 mL Schlenk flask and bubbled with nitrogen for
30 min. The flask was then placed in an oil bath thermostated at
120 °C for 3 h. The polymer was purified by precipitation into
isopropanol cooled with dry ice (Mn=2540; Mw/Mn=1.06).
Experimental Section
Materials. Methyl acrylate (MA, 99%) and styrene (St, 99%)
were purchased from Aldrich and purified by passing through a
column filled with basic alumina to remove inhibitors or anti-
oxidants. Dichloromethane was distilled from calcium hydride.
4-Hydroxy-TEMPO was received from NOVA Molecular Tec-
hnologies Inc. Tris(2-pyridylmethyl)amine (TPMA) was pur-
chased from ATRP Solutions. All other reagents, including (1-
bromoethyl)benzene, 2-bromopropionyl bromide, Cu0 powder
(75 μm), CuBr, CuBr2, N,N,N0,N00,N00-pentamethyldiethylene-
triamine (PMDETA), pyridine, and 2,2,6,6-tetramethylpiper-
idine 1-oxyl (TEMPO), and solvents were purchased from
Aldrich with the highest purity available and used as received
without further purification.
Analyses. NMR spectra were recorded on a Bruker instru-
ment operating at 300 MHz. Monomer conversions were de-
termined by 1H NMR. MW and MWD were determined by size
exclusion chromatography (SEC). The SEC analysis was con-
ducted with a Waters 515 pump and Waters 410 differential
refractometer using Polymer Standards Services (PSS) columns
5
3
2
˚
(Styrogel 10 , 10 , and 10 A) in THF as eluent at 35 °C and at a
flow rate of 1 mL/min. The apparent MW (Mn,SEC and Mw,SEC
)
and Mw/Mn values were determined with a calibration based on
polystyrene (PSt) standards. Mass spectral data were acquired
using a PerSeptive Voyager STR MS matrix-assisted laser
desorption ionization (MALDI) time-of-flight (TOF) mass
spectrometer using the reflection positive ion mode. Dithranol
was used as the matrix, and silver trifluoroacetate as the cation
source. The relaxation delay was set to 450 ns, with an accel-
eration voltage of 20 kV.
Synthesis of the 1-Phenylethylhydroxy-TEMPO Alkoxyamine
(1-(1-Phenylethoxy)-2,2,6,6-tetramethylpiperidin-4-ol), 3. A
Schlenk flask was charged with hydroxy-TEMPO (7.89 g,
45.8 mmol), copper powder (11.6 g, 183 mmol), copper
bromide (52.5 mg, 0.366 mmol), and 50 mL of anisole. The
solution was degassed by bubbling with nitrogen for 30 min,
and then PMDETA (76.6 μL, 0.366 mmol) was added. After
5 min, (1-bromoethyl)benzene (5 mL, 36.6 mmol) was added,

€
Nicolay and Matyjaszewski
242 Macromolecules, Vol. 44, No. 2, 2011
Scheme 2. Synthesis of the 2-Bromopropionate Alkoxyamine Initiator, 4
Table 1. ATRP of MA and St with the 2-Bromopropionate Alkoxyamine 4 and Synthesis of r-Nitroxy-ω-bromo and Cyclic (Co)polymers
entry
M/I/CuBr/CuBr2/L
conv (%)
Mn,th
Mn,SEC; Mw/Mn
Mn,SEC; Mw/Mn after deprotection
Mn,SEC; Mw/Mn after cyclization
1a
2b
3c
72/1/0.5/0.05/0.55
131/1/0.75/0.25/1
262/1/0.75/0.25/1
38
18.8
4.8
2790
2990
3930
2620; 1.09
3200; 1.10
3670; 1.07
2540; 1.06
3030; 1.09
3530; 1.07
1850; 1.07
2580; 1.24
2900; 1.12
a M/I/L = MA/4/PMDETA, MA/acetone = 2/1 v/v, 40 °C, 20.5 h. b M/I/L = St/4/TPMA, St/anisole = 2/1 v/v, 40 °C, 220 min. c M/I/L = St/PMA
(from entry 1)/TPMA, St/anisole/acetone = 2/2/1 v/v/v, 40 °C, 360 min.
General Procedure for the Synthesis of Cyclic PMA, PSt, and
PMA-b-PSt. In a typical experiment, CuBr (28.7 mg, 0.2 mmol),
TPMA (58.1 mg, 0.2 mmol), and Cu0 (31.8 mg, 0.5 mmol) were
charged to a flask. The flask was deoxygenated by purging with
N2 for 30 min, and then deoxygenated acetone (30 mL) was
introduced. The contents of the flask were stirred for 15 min, and
then a solution of R-nitroxy-ω-bromo PMA (2,540, 50.8 mg,
0.02 mmol) in 20 mL of deoxygenated acetone was introduced
over a period of 30 h. The solution was subsequently stirred for
5 h at RT. The solution was passed through a neutral alumina
column to remove the copper catalyst and concentrated under
vacuum, and the resulting polymer was analyzed by SEC with-
out any additional treatment (Mn=1850; Mw/Mn=1.07).
General Procedure for the Ring Expansion of Cyclic PSt by
NMP of St. In a typical experiment, cyclic PSt (22 mg, 0.01
Figure 1. SEC traces of the R-aminooxy-ω-bromo PMA, the R-nitroxy-
ω-bromo PMA obtained after deprotection of the alkoxyamine function,
mmol, Mn=2200, Mw/Mn=1.24) and deoxygenated St (0.6 mL,
5.24 mmol) were charged in a deoxygenated flask. The flask was
placed in an oil bath thermostated at 125 °C for 10 h. The
polymerization was stopped via immersion in an ice bath.
and the final cyclic PMA.
Deprotection of the nitroxide function was achieved by
heating the R-aminooxy-ω-bromo PMA in anisole ([PMA]=
26.7 mmol/L), at 120 °C for 3 h, in the presence of 100 equiv
of TEMPO. The deprotected PMA (Mn = 2540 and Mw/
Mn=1.06) was recovered by precipitation into cold (dry ice)
isopropanol (Figure 1). The fact that deprotection of the
nitroxide function had occurred was confirmed by MALDI-
TOF analysis. A loss of mass 107.7 g/mol, corresponding
to the terminal phenylethyl group plus a proton, was
observed going from the R-aminooxy-ω-bromo PMA
population to the R-nitroxy-ω-bromo PMA population
(Figure 2). The isotopic pattern of the bromine atom could
not be observed in the spectra, indicating that the bromine
chain-end of the polymers was cleaved during the ioniza-
tion process.
The telechelic PMA was subsequently cyclized by ATRC
at RT under nitrogen atmosphere and high dilution condi-
tions. A 20 mL deoxygenated solution of R-nitroxy-ω-bromo
PMA (n=0.02 mmol, Mn=2540, Mw/Mn=1.06) in acetone
was introduced over a period of 30 h (0.67 mL/h) into a
30 mL deoxygenated solution of acetone containing 0.2 mmol
of CuBr/TPMA complex and 0.5 mmol of Cu0. After com-
pletion of the addition of the R-nitroxy-ω-bromo PMA
solution, the resulting reaction mixture was stirred for an
additional 5 h at RT. The SEC analysis of the polymer after
reaction showed a decrease of the apparent molecular weight
(Mn=1850; Mw/Mn=1.07) (Figure 1).
Results and Discussion
Synthesis of the 2-Bromopropionate Alkoxyamine Initiator,
4. An ATRP initiator carrying alkoxyamine and bromoester
groups was prepared in two steps (Scheme 2). The first step
was the synthesis of the 1-phenylethylhydroxy-TEMPO
alkoxyamine (1-(1-phenyl-ethoxy)-2,2,6,6-tetramethylpiper-
idin-4-ol), 3, in 80.2% yield, by an ATRC reaction between
(2-bromoethyl)benzene, 1, and 4-hydroxy-TEMPO, 2. The
2-bromopropionate alkoxyamine, 4, was obtained, in quan-
titative yield by condensation of 3 with 2-bromopropionyl
bromide.
Synthesis of Cyclic PMA. The conditions for the ATRP of
MA with the 2-bromopropionate alkoxyamine 4 were cho-
sen to limit, as much as possible, termination events between
propagating radicals, since any such reactions would affect
the chain-end functionality of polymer chains. Therefore,
10% of CuII/L deactivator was initially introduced to limit
the concentration of propagating radicals by shifting the
ATRP equilibrium toward the dormant species and reducing
the concentration of radicals in the system. In addition, to
suppress the homolytic cleavage of the terminal alkoxyamine
bond, a low reaction temperature (40 °C) was used, and the
polymerization was stopped at relatively low conversion
(38%) (Table 1).
The ATRP of MA, initiated by the 2-bromopropionate
alkoxyamine 4, was conducted in 33% acetone solution (v/v) at
40 °C with a ratio [MA]0/[4]0/[CuBr]0/[CuBr2]0/[PMDETA]0=
72/1/0.5/0.05/0.55 (Table 1). After 20.5 h, a well-defined
PMA, with Mn = 2620 (Mn,th = 2790) and Mw/Mn = 1.09,
was obtained (Figure 1).
The ratio of the apparent molecular weights of the macro-
cyclic PMA and of its R-nitroxy-ω-bromo linear precursor is
0.73, in excellent agreement with the literature.31,32 This
result confirms the cyclic nature of the polymer obtained
by intramolecular ATRC of the R-nitroxy-ω-bromo PMA,
under high dilution conditions. Furthermore, the SEC trace

Article
Macromolecules, Vol. 44, No. 2, 2011 243
Figure 2. MALDI-TOF spectra of the R-aminooxy-ω-bromo PMA (top) and of the R-nitroxy,ω-bromo PMA (bottom) obtained after deprotection of
the alkoxyamine.
of the final polymer does not show any tailing toward high
molecular weight, indicating an essentially quantitative yield
of the cyclic polymer via intramolecular process.
Synthesis of Cyclic PSt. The conditions for the ATRP of St
were also selected to limit, as much as possible, termination
events between propagating radicals (Table 1).
The ATRP of St, initiated by the 2-bromopropionate
alkoxyamine 4, yielded a well-defined PSt, with Mn =3200
(Mnth=2990) and Mw/Mn=1.10 (Table 1 and Figure 3).
Deprotection of the nitroxide function was achieved by
heating the R-aminooxy-ω-bromopolystyrene in anisole at
120 °C for 3 h, in the presence of 100 equiv of TEMPO. The
deprotected PSt (Mn=3030 and Mw/Mn=1.09) was purified
Figure 3. SEC traces of the R-aminooxy-ω-bromo PSt, the R-nitroxy-
ω-bromo PSt obtained after deprotection of the alkoxyamine function,
and the final cyclic PSt.
by three successive precipitations into cold (dry ice) isopro-
panol (Table 1).
The telechelic PSt was subsequently cyclized by ATRC at
RT under nitrogen atmosphere and high dilution conditions.
The SEC analyses of the polymers after reaction demonstrate
a decrease of the apparent molecular weight, with Mn=2580
(Mw/Mn = 1.24) (Table 1). The SEC trace of the final PSt
shows the presence of higher molecular weight polymer,
indicating that some intermolecular coupling occurred dur-
ing the cyclization (estimated to be ∼5% by deconvolution of
the SEC trace) (Figure 3). The occurrence of intermolecular
coupling under these conditions could indicate that the linear
precursor was not perfectly functionalized. Chain-end func-
tionality defects could result from the partial homolytic
cleavage of the terminal C-Br bond during the deprotec-
tion of the R-aminooxy-ω-bromopolystyrene at 120 °C.
However, the ratio of the apparent molecular weights of
the final PSt and of the R-nitroxy-ω-bromo linear precursor
is 0.85, demonstrating that most of the final polymer chains
were the product of intramolecular cyclization.31,32
Synthesis of Cyclic PMA-b-PSt. Diblock copolymers de-
monstrate a unique ability to self-assemble into complex
nanoscale morphologies in bulk (e.g., lamellae, hexagonally
packed cylinders, bicontinuous gyroids, and body-centered-
cubic arrays of spheres) or in solution (e.g., micelle, vesicle,
tube).2,33-36 Cyclic block copolymers could self-assemble in
new morphologies in solution or in bulk due to their unique
topology.2,36

€
Nicolay and Matyjaszewski
244 Macromolecules, Vol. 44, No. 2, 2011
Cyclic PMA-b-PSt block copolymers were prepared in
order to exemplify the wide range of cyclic polymers acces-
sible via this new approach. A linear R-aminooxy-ω-bromo
PMA (entry 1, Table 1), prepared with the 2-bromopropio-
nate alkoxyamine 4, was chain extended by ATRP of St to
shows the presence of higher MW copolymer, indicating that
a small fraction of the copolymer underwent intermolecular
coupling during the cyclization reaction (estimated to be less
than 2% by deconvolution of the SEC trace) (Figure 4).
Ring Expansion of Cyclic PSt. The direct synthesis of well-
defined cyclic polymers from cyclic initiators is an attractive
route to cyclic polymers as it eliminates the need to prepare
linear precursors. However, this approach requires cyclic
initiators and polymerization processes in which chain trans-
fer and exchange reactions between active species are absent.
There have been only three reports in the literature regarding
the synthesis of macrocycles by ring expansion of cyclic
alkoxyamines.28,37,38 These reports indicated difficulty asso-
ciated with the preparation of well-defined macrocycles by
NMP from cyclic alkoxyamines. Therefore, we were inter-
ested in conducting further studies on this process using
macrocyclic alkoxyamines prepared by ATRC.
A macrocyclic polystyrene alkoxyamine (Mn=2200; Mw/
Mn = 1.24) was used as macroinitiator for the NMP of
St. The polymerization was conducted in bulk over 10 h at
125 °C with an initial ratio [St]/[MI]=525 (Figure 5 and
Table 2). SEC analyses of the reaction medium during the
NMP of St from the macrocyclic polystyrene alkoxyamine
revealed the existence of two populations (Figure 5). A high
MW population corresponding to a mixture of cyclic and
linear polystyrene chains, with approximately three to four
alkoxyamine functions per chain on average (Table 2). The
high MW population peak shifts toward higher MW as the
reaction progressed (Figure 5 and Table 2). In contrast, the
peak corresponding to a low MW population does not shift
as the reaction progressed and its relative intensity decreases
with time (Figure 5 and Table 2). The low MW population
corresponds to monocyclic PSt and to dead chains. Decon-
volution of the final SEC curve (Figure 5, 10 h) indicates that
the low and high MW populations represent approximately
22% and 78% of the total number of chains, respectively.
This corresponds to a weight fraction of 1% for the low MW
population and 99% for the high MW population.
yield a well-defined PMA-b-PSt, with Mn = 3930 (Mn,th
=
3670) and Mw/Mn = 1.07 (entry 3, Table 1). The SEC
chromatogram of the linear PMA-b-PSt block copolymer
did not present any shoulder toward low molecular weight,
indicating high initiation efficiency for the block extension
from the PMA macroinitiator (MI) (Figure 4). Deprotection
and cyclization of the resulting R-nitroxy-ω-bromo PMA-b-
PSt copolymer were conducted under the same conditions as
described for the homopolymers. The apparent MW of the
cyclic block copolymer (Mn = 2900; Mw/Mn = 1.12) was
significantly lower (Mn=3530; Mw/Mn=1.07) than that of
its linear R-nitroxy-ω-bromo PMA-b-PSt precursor, con-
firming the efficient cyclization (Figure 4 and entry 3,
Table 1). The SEC trace of the final PMA-b-PSt copolymer
Figure 4. SEC traces of the R-aminooxy-ω-bromo PMA, the R-amino-
oxy-ω-bromo PMA-b-PSt, the R-nitroxy-ω-bromo PMA-b-PSt obtained
after deprotection of the alkoxyamine function, and the final cyclic
PMA-b-PSt.
These results are in good agreement with previous litera-
ture reports and confirm the challenges for ring expansion of
cyclic (macro)alkoxyamines.28,37,38
Three phenomena, namely the PRE,39,40 radical crossover
reactions,24,41 and styrene thermal self-initiation,42,43 are re-
sponsible for the fact that ring expansion of (macro)cyclic
alkoxyamine did not yield well-defined (mono)macrocycles
but a mixture of cyclic and linear polyadducts. While styrene
thermal self-initiation is specific to styrene, the PRE and
radical crossover reaction are inherent to any NMP systems.
The persistent radical effect in NMP describes the variation of
concentrations of growing (transient) and persistent (nitroxyl)
radicals involved in this polymerization process.39,40,44 At the
early stages, a (small) fraction of the growing radicals R•
formed by dissociation of the unimolecular cyclic macro-
initiator undergo radical-radical coupling. This leads, in the
present case, to terminated PSt (•Y;RR;Y• species) and
loss of two growing radicals, R• (Scheme 3). However, by its
nature, the persistent radical, Y•, does not undergo coupling,
Figure 5. Evolution of SEC traces during the NMP of St with a cyclic
PSt MI. Conditions: St/MI=525, bulk, 125 °C, 10 h.
Table 2. Ring Expansion of Cyclic PSt by NMP of Sta
time conv ln([M]0/
(h) (%) [M]t)
Mn,SEC(high MW population)
Mn,th
/
Mn,SEC(high MW population)
Mn,th cyclic
/
b
Mn,th Mn,th cyclic Mn,SEC(low MW population) Mn,SEC(high MW population)
c
0
2
5
0
0
2 200
5 750
13 600
19 400
1 650
4 310
10 200
14 500
6.5 0.0673
20.9 0.234
1950
1920
1970
17 000
45 000
54 300
2.96
3.30
2.80
3.94
4.41
3.73
10 31.5 0.378
a [St]0/[MI]0 = 525, bulk, 125 °C, 10 h. b Mn,th = [St]0/[MI]0 ꢀ 104.15 ꢀ conv þ 2200; no correction made regarding the cyclic nature of the polymers.
c Mn,th cyclic corresponds to the theoretical apparent Mn that would be observed by SEC assuming a cyclic topology: Mn,th cyclic=Mn,th ꢀ 0.75.

Article
Macromolecules, Vol. 44, No. 2, 2011 245
and its overall concentration should continuously increase.
This increase of [Y•] is self-controlling since a higher con-
centration leads to more efficient formation of the dor-
mant alkoxyamines R-Y and a decrease in the amount of
radical-radical coupling (Scheme 3). Therefore, the persis-
tent radical effect, which is a key feature to the nature of
control in a NMP, breaks the stoichiometric balance between
growing and persistent radicals (needed for preservation of
cyclic structures).
indicate that only a small fraction of chains were lost
through irreversible termination after the equilibrium
was reached.
However, such a calculation does not take into account the
fact that styrene undergoes spontaneous thermal polymeri-
zation at high temperatures.42,43 It has been shown, for the
bulk polymerization of styrene at 125 °C in the presence of
the PSt-TEMPO adduct, that the rate of polymerization is
independent of the adduct concentration and is equal to the
rate of thermal polymerization.46,48-50 The proposed kinetic
scheme assumes a stationary state with respect to both [R•]
and [Y•], with a concentration of R• determined by the
equality of the rate of thermal initiation and that of R•-R•
biradical termination.46,49-51 Under the conditions used in
our study for the ring expansion of cyclic PSt, i.e., 10 h in bulk
at 125 °C, the concentration of radicals generated during the
course of the reaction could be as much as 5 mM, that is, ∼30%
of the initial concentration of cyclic PSt.52
The unbalanced stoichiometry between growing and per-
sistent radicals, resulting from the PRE (Scheme 4a), and the
generation of new chains by thermal initiation (Scheme 4b)
both shift the topology of polymer chains from purely cyclic
to a mixture of linear and cyclic polymers (Scheme 4).
This topological change also reflects radical crossover
reactions41 taking place during NMP. Radical crossover
corresponds to the recombination of a growing radical R1
with a persistent radical Y2 generated from the homolytic
cleavage of two distinct alkoxyamines R1-Y1 and R2-Y2.
For example, it has beenreported that the NMP of St at 125°C,
using a 1:1 mixture of unfunctionalized alkoxyamine, 5, and
dihydroxy alkoxyamine, 6, yielded a mixture of polystyrene
macromolecules where all four possible products (7, 8, 9,
and 10) were present in an approximately statistical ratio
(Scheme 5).41 It was concluded from these results that
persistent nitroxide radicals are free to diffuse out of the
reaction cage during NMP and that statistical exchange of
the chain ends should occur even at an early stage of the
polymerization.
The equilibrium concentration of propagating radicals
can be easily calculated from the rate of polymerization
(eq 1). This concentration was found to be equal to 4.55 ꢀ
10-9 mol/L, using a propagation rate constant of 2310 L
mol-1 s-1 (pre-exponential factor of 107.63 L mol s-l and
activation energy of 32.5 kJ mol
)
-1 45,46 for styrene at 125 °C.
ꢀ
ꢁ
- d½Mꢁ
½Mꢁ0
½Mꢁ
Rp
¼
¼ kp½R•ꢁ½Mꢁ w ln
dt
ꢀ
ꢁ
½Mꢁ0
½Mꢁ
ln
¼ kp½R•ꢁt w ½R•ꢁ ¼
kpt
¼ 4:55 ꢀ 10- 9 mol=L
ð1Þ
Knowing the concentration of propagating radicals at the
equilibrium, one can estimate the concentration of chains
that underwent termination after the equilibrium was reached
(eq 2).
- d½R•ꢁ
dt
d½Pꢁ
dt
2
2
Rt ¼
¼
¼ 2kt½R•ꢁ w ½Pꢁ ¼ 2kt½R•ꢁ t
¼ 5:96 ꢀ 10- 4 mol=L
ð2Þ
Taking a rate constant of termination of ∼4 ꢀ 108 L
mol-1 s-1 for styrene at 125 °C,47 the estimated concentra-
tion of dead chains after 10 h should be 5.96 ꢀ 10-4 mol/L,
i.e., ∼3.6% of the initial MI concentration. This would
Crossover reactions between growing and persistent radi-
cals are incompatible with the formation of well-defined
macrocycles. These exchange reactions will lead to the for-
mation of different populations that could still present a
cyclic topology but will contain different number of alkoxy-
amine functions per chain (Scheme 6). Such reshuffling
reactions will occur even if the stoichiometric balance be-
tween growing and persistent radicals is preserved. In addi-
tion, in the presence of persistent radicals and/or thermal
Scheme 3. General Scheme for NMP from Cyclic Alkoxyamines
Scheme 4. Equilibrium between Linear and Cyclic Polymer Chains during NMP with Cyclic Alkoxyamines

€
Nicolay and Matyjaszewski
246 Macromolecules, Vol. 44, No. 2, 2011
Scheme 5. Radical Crossover during NMP of St from a 1:1 Mixture of Unfunctionalized and Dihydroxy Alkoxyamines41
Scheme 6. Population Redistribution via Radical Crossover during NMP with Cyclic Alkoxyamines
initiation, crossover reactions will convert part of the cyclic
polymers into linear polyadducts (Scheme 4).
(3) De Gennes, P. G. J. Chem. Phys. 1971, 55, 572–579.
(4) Sperling, L. H. In Introduction to Physical Polymer Science, 2nd ed.;
Wiley: New York, 1992; pp 104-105.
(5) McLeish, T. C. B. Adv. Phys. 2002, 51, 1379–1527.
(6) Bielawski, C. W.; Benitez, D.; Grubbs, R. H. Science 2002, 297,
2041–2044.
(7) Bielawski, C. W.; Benitez, D.; Grubbs, R. H. J. Am. Chem. Soc.
2003, 125, 8424–8425.
(8) Boydston, A. J.; Holcombe, T. W.; Unruh, D. A.; Frechet, J. M. J.;
Grubbs, R. H. J. Am. Chem. Soc. 2009, 131, 5388–5389.
(9) Braunecker, W. A.; Matyjaszewski, K. Prog. Polym. Sci. 2007, 32,
93–146.
(10) Wang, J.-S.; Matyjaszewski, K. J. Am. Chem. Soc. 1995, 117, 5614–
5615.
(11) di Lena, F.; Matyjaszewski, K. Prog. Polym. Sci. 2010, 35, 959–
1021.
(12) Matyjaszewski, K.; Xia, J. Chem. Rev. 2001, 101, 2921–2990.
(13) Matyjaszewski, K.; Tsarevsky, N. V. Nature Chem. 2009, 1,
276–288.
(14) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int. Ed.
2001, 40, 2004–2021.
(15) Tsarevsky, N. V.; Sumerlin, B. S.; Matyjaszewski, K. Macromole-
cules 2005, 38, 3558–3561.
(16) Binder, W. H.; Sachsenhofer, R. Macromol. Rapid Commun. 2007,
28, 15–54.
(17) Laurent, B. A.; Grayson, S. M. J. Am. Chem. Soc. 2006, 128, 4238–
4239.
(18) Coessens, V.; Matyjaszewski, K. J. Macromol. Sci., Pure Appl.
Chem. 1999, A36, 667–679.
(19) Xu, J.; Ye, J.; Liu, S. Macromolecules 2007, 40, 9103–9110.
(20) Eugene, D. M.; Grayson, S. M. Macromolecules 2008, 41, 5082–
5084.
Conclusion
A novel approach to prepare cyclic polymers via a combination
of ATRP and ATRC was developed. A functional ATRP initiator
carrying an alkoxyamine group was synthesized in two steps with
an overall yield of 80%. This initiator was successfully employed to
prepared well-defined linear telechelic R-nitroxy,ω-bromo homo-
and block copolymers via ATRP of MA or St and subsequent
thermal deprotection of the terminal nitroxide function. Cycliza-
tion reactions were achieved by intramolecular ATRC, under high
dilution conditions. Well-defined, low MW cyclic PMA, PSt as
well as cyclic PMA-b-PSt block copolymers were prepared using
this approach. This approach requires a minimal number of high
yield steps to prepare well-defined macrocycles by radical means.
The ring expansion of a cyclic polystyrene alkoxyamine prepared
by ATRC was conducted via NMP of St and resulted in the
formation of a mixture of products, comprising cyclic and linear
polymers with an average of three to four alkoxyamine functions
per chain. The presence of linear polymer chains reflects the
nonstoichiometric balance between growing and persistent radi-
cals, caused by the PRE, as well as the generation of new chains by
thermal initiation. Radical crossover amplifies these phenomena
and also leads to the formation of cyclic polycondensates during
NMP. Therefore, ring expansion of cyclic (macro)alkoxyamines
by NMP does not appear to be a suitable approach, by nature, for
the preparation of well-defined macrocycles.
(21) Chiefari, J.; Chong, Y. K.; Ercole, F.; Krstina, J.; Jeffery, J.; Le,
T. P. T.; Mayadunne, R. T. A.; Meijs, G. F.; Moad, C. L.; Moad,
G.; Rizzardo, E.; Thang, S. H. Macromolecules 1998, 31, 5559–
5562.
(22) Moad, G.; Rizzardo, E.; Thang, S. H. Aust. J. Chem. 2006, 59,
669–692.
(23) Georges, M. K.; Veregin, R. P. N.; Kazmaier, P. M.; Hamer, G. K.
Acknowledgment. The authors thank Professor JPA Heuts
for valuable discussions. The authors thank the NSF (CHE-07-
15494 and DMR-0549353) and the members of the CRP Con-
sortium at Carnegie Mellon University for their financial sup-
€
port. R. Nicolay thanks Dr. and Mrs. McWilliams (Astrid and
Bruce McWilliams Fellowship) for their financial support.
Macromolecules 1993, 26, 2987–2988.
(24) Hawker, C. J.; Bosman, A. W.; Harth, E. Chem. Rev. 2001, 101,
3661–3688.
(25) Qiu, X.-P.; Tanaka, F.; Winnik, F. M. Macromolecules 2007, 40,
7069–7071.
(26) Goldmann, A. S.; Quemener, D.; Millard, P.-E.; Davis, T. P.;
Stenzel, M. H.; Barner-Kowollik, C.; Mueller, A. H. E. Polymer
2008, 49, 2274–2281.
(27) O’Bryan, G.; Ningnuek, N.; Braslau, R. Polymer 2008, 49, 5241–
5248.
References and Notes
(1) Semlyen, J. A., Ed.; Cyclic Polymers, 2nd ed.; Kluwer Academic
Publishers: Dordrecht, 2000; p 790.
(2) Deffieux, A.; Borsali, R. Controlled synthesis and properties of
cyclic polymers. In Macromolecular Enginering; Matyjaszewski, K.,
Gnanou, Y., Leibler, L., Eds.; Weinheim: Germany, 2007; Vol. 2,
pp 875-908.

Article
Macromolecules, Vol. 44, No. 2, 2011 247
(28) Narumi, A.; Zeidler, S.; Barqawi, H.; Enders, C.; Binder, W. H.
J. Polym. Sci., Part A: Polym. Chem. 2010, 48, 3402–3416.
(29) Matyjaszewski, K.; Woodworth, B. E.; Zhang, X.; Gaynor, S. G.;
Metzner, Z. Macromolecules 1998, 31, 5955–5957.
(30) Nicolay, R.; Marx, L.; Hemery, P.; Matyjaszewski, K. Macro-
molecules 2007, 40, 9217–9223.
(41) Hawker, C. J.; Barclay, G. G.; Dao, J. J. Am. Chem. Soc. 1996, 118,
11467–11471.
(42) Hui, A. W.; Hamielec, A. E. J. Appl. Polym. Sci. 1972, 16, 749–769.
(43) Mardare, D.; Matyjaszewski, K. Polym. Prepr. (Am. Chem. Soc.,
Div. Polym. Chem.) 1994, 35, 778–779.
(44) Tang, W.; Fukuda, T.; Matyjaszewski, K. Macromolecules 2006,
39, 4332–4337.
€
(31) Keul, H.; Hocker, H. Cycloalkanes and Related Oligomers and
Polymers. In Large Ring Molecules; Semlyen, J. A., Ed.; Wiley:
New York, 1996; p 397.
(32) Lepoittevin, B.; Perrot, X.; Masure, M.; Hemery, P. Macromole-
(45) Manders, B. G.; Chambard, G.; Kingma, W. J.; Klumperman, B.;
van Herk, A. M.; German, A. L. J. Polym. Sci., Part A: Polym.
Chem. 1996, 34, 2473–2479.
cules 2001, 34, 425–429.
(33) Bates, F. S. Science 1991, 251, 898–905.
(46) Fukuda, T.; Terauchi, T. Chem. Lett. 1996, 293–294.
(47) Buback, M.; Busch, M.; Kowollik, C. Macromol. Theory Simul.
2000, 9, 442–452.
(48) Greszta, D.; Matyjaszewski, K. Macromolecules 1996, 29, 5239–
5240.
(49) Fukuda, T.; Terauchi, T.; Goto, A.; Ohno, K.; Tsujii, Y.; Miyamoto,
T.; Kobatake, S.; Yamada, B. Macromolecules 1996, 29, 6393–
6398.
(34) Lodge, T. P. Macromol. Chem. Phys. 2003, 204, 265–273.
(35) Lecommandoux, S.; Borsali, R. Polym. Int. 2006, 55, 1161–1168.
(36) Schappacher, M.; Deffieux, A. Science 2008, 319, 1512–1515.
(37) Lepoittevin, B. Syntheꢀse et caracteꢁrisation d’architectures aꢀ base de
polystyreꢀne cyclique preꢁpareꢁ par polymeꢁrisations contro^leꢁes; Pierre
et Marie Curie University (Paris 6): Paris, 2000.
(38) Ruehl, J.; Ningnuek, N.; Thongpaisanwong, T.; Braslau, R.
J. Polym. Sci., Part A: Polym. Chem. 2008, 46, 8049–8069.
(39) Fischer, H. Macromolecules 1997, 30, 5666–5672.
(40) Fischer, H. Chem. Rev. 2001, 101, 3581–3610.
(50) Greszta, D.; Matyjaszewski, K. Macromolecules 1996, 29, 7661–
7670.
(51) Goto, A.; Fukuda, T. Prog. Polym. Sci. 2004, 29, 329–385.
(52) Tobolsky, A. V.; Baysal, B. J. Polym. Sci. 1953, 11, 471–486.