Synthesis of hyperbranched glycopolymers via self-condensing atom transfer radical copolymerization of a sugar-carrying acrylate

Article abstract of DOI:10.1021/ma048436c

Hyperbranched glycopolymers were synthesized by self-condensing vinyl copolymerization (SCVCP) of an acrylic AB* inimer, 2-(2-bromopropionyloxy)ethyl acrylate (BPEA), with 3-O-acryloyl-1,2: 5,6-di-O-isopropylidene-α-D-glucofuranoside (AIGlc) via atom transfer radical polymerization (ATRP), followed by deprotection of the isopropylidene protecting groups. Homopolymerization of AIGlc with the CuBr/pentamethyldiethylenetriamine (PMDETA) catalyst system in solution resulted in linear poly(AIGlc) having controlled molecular weights and narrow molecular weight distribution, which were characterized using GPC, GPC/viscosity, and MALDI-TOF mass spectrometry. The catalyst system could be applied for SCVCP to synthesize hyperbranched poly(AIGlc)s, in which the molecular weights, the composition of AIGlc segment, and the branched structures can be adjusted by an appropriate choice of the comonomer ratio, γ. Deprotection of the isopropylidene protecting groups of the branched poly(AIGlc)s resulted in water-soluble glycopolymers with randomly branched architectures.

Full text of DOI:10.1021/ma048436c

Macromolecules 2005, 38, 9-18
9
Articles
Synthesis of Hyperbranched Glycopolymers via Self-Condensing Atom
Transfer Radical Copolymerization of a Sugar-Carrying Acrylate
Sharmila Muthukrishnan, Gu1 nter Jutz, Xavier Andre´, Hideharu Mori,† and
Axel H. E. Mu1 ller*
Makromolekulare Chemie II and Bayreuther Zentrum fu¨r Kolloide und Grenzfla¨chen, Universita¨t
Bayreuth, D-95440 Bayreuth, Germany
Received July 29, 2004; Revised Manuscript Received October 13, 2004
ABSTRACT: Hyperbranched glycopolymers were synthesized by self-condensing vinyl copolymerization
(SCVCP) of an acrylic AB* inimer, 2-(2-bromopropionyloxy)ethyl acrylate (BPEA), with 3-O-acryloyl-1,2:
5,6-di-O-isopropylidene-R-D-glucofuranoside (AIGlc) via atom transfer radical polymerization (ATRP),
followed by deprotection of the isopropylidene protecting groups. Homopolymerization of AIGlc with the
CuBr/pentamethyldiethylenetriamine (PMDETA) catalyst system in solution resulted in linear poly(AIGlc)
having controlled molecular weights and narrow molecular weight distribution, which were characterized
using GPC, GPC/viscosity, and MALDI-TOF mass spectrometry. The catalyst system could be applied
for SCVCP to synthesize hyperbranched poly(AIGlc)s, in which the molecular weights, the composition
of AIGlc segment, and the branched structures can be adjusted by an appropriate choice of the comonomer
ratio, γ. Deprotection of the isopropylidene protecting groups of the branched poly(AIGlc)s resulted in
water-soluble glycopolymers with randomly branched architectures.
Introduction
and hydrophilicity, sugar-based hydrogels are used in
biomedical engineering as superabsorbents, contact
lenses, and matrices for drug delivery systems.10,11
Synthetic carbohydrate-based polymers are now being
used as very important tools to investigate carbohydrate-
based interactions.1,2 Carbohydrates are involved in
various biological functions in living systems. For
instance, heparin, which is a natural polyanion com-
posed of repeating dissaccharide units, is a first polysac-
charide applied in medicine and plays an important role
in blood coagulation.3 Synthetic carbohydrate polymers
with biocompatible and biodegradable properties are
used in tissue engineering and controlled drug release
devices. Many of them are used also as surfactants4 and
biologically active polymers.5
The new era of “glycomimics”, which are synthetic
complex carbohydrates and carbohydrate-based poly-
mers, are increasingly used for investigating glycopoly-
mer-protein interactions.6 These glycopolymers are
widely investigated for pharmaceutical and medical
applications in the treatment of infectious diseases.7
Glycopolymers generally include natural as well as
synthetic carbohydrate-bearing polymers, whereas syn-
thetic polymers containing sugar moieties as pendant
groups can be referred in a narrower sense as glyco-
polymers.8 There are various types of synthetic sugar-
carrying polymers. Linear polymers, comb-shaped poly-
mers, dendrimers, and cross-linked hydrogels represent
the four major classes.9 Because of their biocompatibility
Highly branched glycopolymers have adopted nature’s
multivalent approach to their work,12 and therefore they
have been used to understand the multivalent processes.
It has been reported that polyvalent saccharide ligands
inhibit the protein-carbohydrate interaction, providing
the best therapeutic strategy for the treatment of
myriad human diseases.13 Silylated glycodendrimers
have been tested in bacterial adhesion hemagglutination
assays.14 Dendrimers with covalently attached glycoside
residues in the outer layer are well-defined glycopolymer
models of cell surface multiantennary glycoproteins.15
The polyvalency inherent in carbohydrate-based poly-
mers, especially in branched polymers, is an important
feature which allows these materials to serve as cell
surface mimics to understand and manipulate carbo-
hydrate-protein interactions.
The recent developments in the field of glycoscience,
glycotechnology, and the potential applications of gly-
copolymers have attracted researchers to develop new
synthetic routes to design a variety of sugar-containing
polymers with controlled architectures and functional-
ities. Various polymerization techniques have been
developed to synthesize sugar-containing polymers with
well-defined structures, which involve cationic polym-
erization,16 ring-opening polymerization,17 ring-opening
metathesis polymerization,18 and free radical polymer-
ization.19 Although a variety of glycopolymers has been
synthesized by conventional free radical polymerization
of vinyl monomers carrying sugar residues, it was
difficult to control molecular weights and architecture.
† Present address: Department of Polymer Science and Engi-
neering, Faculty of Engineering, Yamagata University, 4-3-16,
Jonan, Yonezawa, 992-8510, Japan.
* To whom correspondence should be addressed: e-mail
Axel.Mueller@uni-bayreuth.de; phone +49 (921) 55-3399; fax
+49 (921) 55-3393.
10.1021/ma048436c CCC: $30.25 © 2005 American Chemical Society
Published on Web 12/04/2004

10 Muthukrishnan et al.
Macromolecules, Vol. 38, No. 1, 2005
Scheme 1. General Route to Branched Glycopolymers via Self-Condensing Vinyl Copolymerization, Followed by
Deprotection of Isopropylidene Protecting Groups
Controlled/“living” radical polymerization has allowed
well-defined and controlled synthesis of glycopolymers
by a very facile and simple approach. For example,
Fukuda et al. have reported the synthesis of styryl and
methacryloyl monomers having a saccharide residue by
using nitroxide-mediated polymerization20 and atom
transfer radical polymerization (ATRP).21 They also
have reported the synthesis of glycopolymers by nitrox-
ide-mediated polymerization of a sugar-carrying acry-
late, but the system gave polymers with limited con-
version.22 Gotz et al. have reported the synthesis of new
lipo-glycopolymer amphiphiles, which were prepared by
nitroxide-mediated polymerization with a dioctadecyl-
substituted nitroxide initiator and 1,2,5,6-di(isopropy-
lidene)-D-glucose-2-propenoate as monomer.23 ATRP has
been used for the synthesis of many sugar-carrying
block24,25 and graft polymers.26 Very recently, Narain
and Armes reported the first ATRP of unprotected
sugar-based monomers.27,28
This paper reports the synthesis of randomly branched
glycopolymers using self-condensing vinyl copolymeri-
zation (SCVCP) of an acrylic AB* inimer, 2-(2-bro-
mopropionyloxy)ethyl acrylate (BPEA) with 3-O-acryloyl-
1,2:5,6-di-O-isopropylidene-R-D-glucofuranoside (AIGlc)
via ATRP. The synthetic route to highly branched
glycopolymers is given in Scheme 1. The curved lines
represent polymer chains. A*, B*, and M* are active
units, whereas a, b, and m are reacted ones. A is an
acryloyl group. M and m stand for AIGlc units at the
chain end and in the linear segment, respectively.
SCVCP of AB* inimers with conventional monomers is
a facile approach to obtain functional branched polymers
because different types of functional groups can be
incorporated into a polymer, depending on the chemical
nature of the comonomer.29-32 In this study, an isopro-
pylidene-protected sugar-carrying acrylate, AIGlc, was
selected as a comonomer for the synthesis of water-
soluble highly branched glycopolymers. For an ideal
SCVCP process, living polymerization systems are
required to avoid cross-linking reactions and gelation
due to chain transfer or recombination reactions. Hence,
the important step is to find suitable conditions where
both homoplymerization of AIGlc and homo-SCVP of the
inimer can proceed in a controlled/“living” fashion. In
general, Cu-based ATRP was employed for SCV(C)P of
acrylate-type inimers having an acrylate (A) and a
bromoester group (B*), capable to initiate ATRP.33 In
this study, we used the CuBr/pentamethyldiethylen-
etriamine (PMDETA) catalyst system to prepare well-
defined and monodisperse linear poly(AIGlc). Effects of
temperature and solvents were investigated in terms
of the polydispersity and polymerization rate. Randomly
branched poly(AIGlc)s having different molecular weights
and degree of branching were synthesized by SCVCP
with different comonomer ratios, γ, via ATRP.
Experimental Section
Materials. CuBr (95%, Aldrich) was purified by stirring
overnight in acetic acid. After filtration, it was washed with
ethanol, ether, and then dried. N,N,N′,N′′,N′′-Pentamethyldi-
ethylenetriamine (PMDETA, 99%, Aldrich) and ethyl 2-bromo-
2-isobutyrate (98%, Aldrich) were distilled and degassed.
Synthesis of an acrylic AB* inimer, 2-(2-bromopropionyloxy)-
ethyl acrylate (BPEA), was conducted by the reaction of
2-bromo-propinoyl bromide with 2-hydroxyethyl acrylate in the
presence of pyridine as reported previously.34,35 The inimer was
degassed by three freeze-thaw cycles. Other reagents were
commercially obtained and used without further purification.
Synthesis of 3-O-Acryloyl-1,2:5,6-di-O-isopropylidene-
r-D-glucofuranoside (AIGlc). The synthesis of AIGlc was
carried out according to the method reported by Fukuda et
al.22 and Ouchi et al.36 with slight modifications. To a solution
of 1,2:5,6-di-O-isopropylidene-D-glucofuranose (10 g, 38.5 mmol)
in acetone (50 mL) was added a 5 N solution of NaOH (11.55
mL, 57.8 mmol), and then acryloyl chloride (3.4 mL, 38.5
mmol) was added dropwise to the mixture with stirring in an
ice bath. After 1 h it was allowed to stir at room temperature
for 48 h. Then the reaction mixture was diluted with 100 mL
of water and extracted with n-hexane. The aqueous layer was
washed two times with hexane, and the combined extract was
washed three times with water to remove the salts. It was then
dried over anhydrous sodium sulfate. The drying agent was
filtered off, and the filtrate was evaporated under reduced
pressure to give a crude syrup. The crude syrup was distilled
under a very high vacuum (90 °C, 1 × 10-2 mbar) to give white
1
solid, 5.0 g (41%). H NMR (CDCl3): δ ) 1.25-1.55 (m, 12H,

Macromolecules, Vol. 38, No. 1, 2005
Hyperbranched Glycopolymers 11
Table 1. Homopolymerization of 3-O-Acryloyl-1,2:5,6-di-O-isopropylidene-r-D-Glucofuranoside (AIGlc) with CuBr/
PMDETA at 60 °C in Ethyl Acetate (50 wt %)a
c
[M]0/[I]0
time (h)
convb (%)
Mn, calcd
Mn,GPCd (Mw/Mn)
Mn,GPC-VISCOe (Mw/Mn)
Mn, MALDIf (Mw/Mn)
15
20
18
48
73
88
93
33
84
3400
5500
14600
10300
26300
4100 (1.05)
5900 (1.09)
13400 (1.09)
9300 (1.06)
24200 (1.14)
3200 (1.14)
5400 (1.09)
13600 (1.11)
9400 (1.13)
29200 (1.12)
6600 (1.13)
18500 (1.25)
14800 (1.17)
30900 (1.37)
50
72
100
100
48
120
a Solution polymerization with ethyl 2-bromoisobutyrate; [I]0:[CuBr]0:[PMDETA]0 ) 1:1:1. b Monomer conversion as determined by 1H
NMR. c Theoretical number-average molecular weight as calculated from the monomer conversion. d Determined by GPC using THF as
eluent with PtBuA standards. e Determined by GPC-viscosity measurement. f Determined by MALDI-TOF MS measurement.
isopropylidene units), 4.05, 4.25, 5.29, 5.87 (7H, sugar moiety),
5.85-6.50 (3H, three vinyl protons). 13C NMR (CDCl3): δ )
165.05 (CdO), 131.6 (CH2)), 128.09 (CH)), 27.18, 27.07,
26.56, 25.59 (CH3 in isopropylidene groups).
white powder in a quantitative yield (75 mg, yield ) 88%),
which was insoluble in water, methanol, and acetone and
completely soluble in DMSO. The deprotection was done in a
similar way in the case of linear poly(AIGlc), which gave water-
soluble glycopolymers.
Atom Transfer Radical Polymerization of AIGlc. All
polymerizations were carried out in a round-bottom flask
sealed with a plastic cap. A representative example is as
follows: Ethyl 2-bromoisobutyrate (0.0254 g, 0.127 mmol) was
added to a round-bottom flask containing CuBr(I) (0.0178 g,
0.127 mmol), PMDETA (0.0219 g, 0.127 mmol), and AIGlc (0.80
g, 2.54 mmol) in ethyl acetate (0.80 g, 50 wt % to AIGlc). As
soon as the initiator was added to the mixture, the color
changed into green, indicating the start of the polymerization.
The flask was placed in an oil bath at 60 °C for 48 h.
Conversion of the double bonds, as detected by 1H NMR, was
88%. The content in the flask was viscous which was dissolved
in THF. The solution was passed through a silica column, and
the polymer was precipitated from THF into hexane. Finally,
the product was freeze-dried from dioxane and dried under
vacuum at room temperature. The polymer had Mn ) 5900
and Mw/Mn ) 1.09 according to conventional GPC, Mn ) 6600
and Mn/Mw ) 1.13 according to GPC/visosity using universal
calibration, and Mn ) 5400 and Mw/Mn ) 1.09 according to
MALDI-TOF MS measurement.
Characterization. The linear and branched polymers
obtained from AIGlc were characterized by conventional GPC
and GPC/viscosity using THF as eluent at a flow rate of 1.0
mL/min at room temperature. A conventional THF-phase GPC
system was used to obtain apparent molecular weights. GPC
system I; column set: 5 µm PSS SDV gel, 102, 103, 104, 105 Å,
30 cm each; detectors: Waters 410 differential refractometer
and Waters photodiode array detector operated at 254 nm.
Narrow PtBuA standards (PSS, Mainz) were used for the
calibration of the column set I. Molecular weights of the
branched polymers were determined by the universal calibra-
tion principle37 using the viscosity module of the PSS-
WinGPC scientific V 6.1 software package. Linear PMMA
standards (PSS, Mainz) were used to construct the universal
calibration curve. GPC system II; column set: 5 µm PSS SDV
gel, 103 Å, 105 Å and 106 Å, 30 cm each; detectors: Shodex
RI-71 refractive index detector; Jasco Uvidec-100-III UV
detector (λ ) 254 nm); Viscotek viscosity detector H 502B. A
1-methyl-2-pyrrolidone (NMP)-phase GPC system was used to
obtain apparent molecular weights of the hydrolyzed polymers.
GPC system III; column set: two PSS GRAM 7 µm, 1000 and
100 Å columns thermostated at 70 °C; detectors: Waters 486
UV detector (λ ) 270 nm), and Bischoff RI-detector 8110. 50
µL of the sample diluted in NMP (containing 0.05 M LiBr) were
injected at a flow rate of 1 mL/min. Linear PS standards were
used for calibration.
MALDI-TOF mass spectrometry was performed on a
Bruker Reflex III instrument equipped with a 337 nm N2 laser
in the reflector mode and 20 kV acceleration voltage. 2,5-
Dihydroxybenzoic acid (Aldrich, 97%) was used as a matrix.
Samples were prepared from THF solution by mixing matrix
(20 mg/mL) and polymer (10 mg/mL) in a ratio of 4:1. The
number-average molecular weights of the polymers were
determined in the linear mode.
1H and 13C NMR spectra were recorded with a Bruker AC-
250. FT-IR spectra were recorded on a Bruker Equinox 55
spectrometer. The elemental analyses were performed by Ilse
Beetz Mikroanalytisches Laboratorium (Kulmbach).
A mixture of linear poly(AIGlc)s with various molecular
weights was used as comparison in the solution viscosity
studies. Molecular weight for this sample: Mn ) 13 800 and
Mw/Mn ) 1.64 (determined by GPC/viscosity using universal
calibration) as shown in Figure S-1 (see Supporting Informa-
tion).
Self-Condensing Vinyl Copolymerization. A represen-
tative example for the copolymerization (γ ) [AIGlc]0/[BPEA]0
) 1) is as follows: BPEA (0.3995 g, 1.592 mmol) was added to
a round-bottom flask containing CuBr(I) (0.004 78 g, 0.0318
mmol), PMDETA (0.005 49 g, 0.0318 mmol), AIGlc (0.5 g, 1.592
mmol), and ethyl acetate (0.5 g, 50 wt % to AIGlc). As soon as
BPEA was added to the mixture, the color changed into green,
indicating the start of the polymerization. The flask was placed
in an oil bath at 60 °C for 4 h. The mixture was completely
solidified after 4 h when the conversion reached a certain level.
Conversion of the double bonds, as detected by 1H NMR, was
90%. After the mixture was dissolved in THF, and was passed
through a silica column, the polymer was precipitated from
THF into hexane. Then the product was freeze-dried from
dioxane and finally dried under vacuum at room temperature
to yield a white powder. The polymer had Mn ) 7800 and Mw/
Mn ) 1.43 according to conventional GPC and Mn ) 9200 and
Mw/Mn ) 2.84 according to GPC/viscosity using universal
calibration. The resulting polymer was soluble in chloroform,
THF, and acetone but insoluble in methanol, hexane, and
water.
Deprotection. Transformation of randomly branched poly-
(AIGlc) into branched poly(3-O-acryloyl-R,â-D-glucopyranoside)
(AGlc) was achieved under mild acidic conditions. The branched
poly(AIGlc) (γ ) 1; 85 mg) was dissolved in 80% formic acid
(10 mL) and stirred for 48 h at room temperature. Then, 4
mL of water was added, and it was stirred for another 3 h.
The solution was dialyzed using Spectra/PorR (MWCO: 1000)
against Millipore water for 2 days. The solution was evapo-
rated under reduced pressure, and the resulting polymer was
freeze-dried from dioxane and dried under vacuum. The
deprotected hyperbranched polymer (γ ) 1) was obtained as
Results and Discussion
Effect of Polymerization Conditions on the Ho-
mopolymerization of AIGlc. To find the suitable
polymerization conditions for the synthesis of highly
branched glycopolymers by SCVCP, we first investi-
gated the effect of polymerization conditions on ATRP
of AIGlc. In a previous paper, we have demonstrated
that the synthesis of randomly branched polymers by
SCVCP of tert-butyl acrylate with BPEA was achieved
using CuBr/PMDETA catalyst system.38 Hence, the
CuBr/PMDETA catalyst system was employed for ATRP
of AIGlc. When AIGlc was polymerized using CuBr/
PMDETA with ethyl 2-bromoisobutyrate ([M]0/[I]0 ) 20)
at 60 °C in ethyl acetate, as shown in Table 1, the
1
conversion reached 88% (as determined by H NMR)

12 Muthukrishnan et al.
Macromolecules, Vol. 38, No. 1, 2005
Table 2. Effects of Solvent and Temperature on Homopolymerization of AIGlc with CuBr/PMDETA
solventa
temp (°C)
[M]0/[I]0
time (h)
convb (%)
Mn,calcd
Mn,GPCc (Mw/Mn)
ethyl acetate
ethyl acetate
ethyl acetate
ethyl acetate
anisole
60
60
50
100
50
100
50
100
72
48
48
24
24
24
93
33
77
47
93
79
14600
10300
12800
14700
15700
23100
13400 (1.09)
9300 (1.06)
14900 (1.27)
13500 (1.18)
14700 (1.25)
24800 (1.26)
80
80
80
anisole
100
a 50 wt %; [I]0:[CuBr]0:[PMDETA]0 ) 1:1:1. b Monomer conversion as determined by 1H NMR. c Determined by GPC using THF as
eluent with PtBuA standards.
Figure 2. GPC traces of linear poly(AIGlc) (a) in ethyl acetate
at 60 and 80 °C ([M]0/[I]0 ) 50) and (b) in anisole at 80 °C
([M]0/[I]0 ) 50) and 100 °C ([M]0/[I]0 ) 100). See Table 2 for
polymerization conditions.
Figure 1. (a) GPC trace (RI signal) and (b) MALDI-TOF
mass spectrum (linear mode) of linear poly(AIGlc) obtained
by CuBr/PMDETA at [M]0/[I]0 ) 20. See Table 1 for detailed
polymerization conditions.
poly(AIGlc)s with low narrow polydispersity (1.05 < Mw/
Mn < 1.14) with predicted molecular weights in the
range of [M]0/[I]0 ) 20-100. These results suggest that
the CuBr/PMDETA catalyst at 60 °C is a suitable
system for the controlled polymerization of AIGlc. In all
cases, the expected molecular weights and narrow
molecular weight distributions could be achieved re-
gardless of [M]0/[I]0, suggesting that the extremely slow
polymerization compared to other type of acrylates is
simply due to the steric hindrance by the bulky side
group in the sugar-carrying acrylate. A similar tendency
was also observed in the nitroxide-mediated polymeri-
zation of AIGlc, which resulted in limited conversion
(∼55%) and molecular weights (Mn ) ∼13 000) with
relatively low polydispersity (1.2 < Mw/Mn < 1.6).22 On
the other hand, our ATRP system provided polymers
with almost full conversion having higher molecular
weights (Mn > 2 × 104) and narrower molecular weight
distributions (Mw/Mn < 1.14).
The homopolymerization of AIGlc was conducted
under various conditions, aiming at increasing the
polymerization rate as well as understanding the effects
of solvent and temperature. The results are summarized
in Table 2. The polymerization in ethyl acetate at 80
°C led to increase of the reaction rate but gave a bimodal
distribution curve, as shown in Figure 2a. Note that the
polymerization at higher concentration (with less sol-
vent) is also not suitable because the solubility of the
monomer is not so high; the solubilization process is
slow even in ethyl acetate (50 wt % to AIGlc) and takes
a prolonged time to dissolve completely. The polymer-
ization in anisole at 80 and 100 °C also gave polymers
having bimodal distribution (Figure 2b). This could be
after 48 h. The number-average molecular weight of
poly(AIGlc) as determined by conventional GPC using
PtBuA standards was Mn ) 5900, which is almost the
same as the theoretical value (Mn ) 5500), and the
polydispersity index was Mw/Mn ) 1.09. As can be seen
later, the linear and branched polymers obtained from
AIGlc were evaluated by conventional GPC and GPC/
viscosity systems, as the relation between molecular
weight and hydrodynamic volume of branched polymers
differs substantially from the linear ones. In addition,
the bulky side group in poly(AIGlc) may lead to different
hydrodynamic volume compared to that of standard
PtBuA. To clarify these points, further characterization
was conducted using GPC/viscosity and MALDI-TOF
mass spectrometry (MS). As can be seen in Figure 1,
the molecular weights and molecular weight distribu-
tions obtained from MALDI-TOF MS (Mn ) 5400 and
Mw/Mn ) 1.09) and GPC/viscosity (Mn ) 6600 and Mw/
Mn ) 1.09) are in agreement with those obtained from
the conventional GPC using PtBuA standards and with
the theoretical values.
Table 1 shows the results of homopolymerization
obtained at different monomer-to-initiator ratios, [M]0/
[I]0. For the cases of the polymerization at [M]0/[I]0 )
100, a longer polymerization time was required to attain
higher conversion (conversion ) 84% for 120 h and 33%
for 48 h). The polymerization mixture was still liquid
after 48 h, but it became viscous and the polymerization
was stopped after 120 h, whereas almost full conversion
was obtained after 72 h in the case of [M]0/[I]0 ) 50. As
can be seen in Table 1, the polymerization in ethyl
acetate using the CuBr/PMDETA system gave linear

Macromolecules, Vol. 38, No. 1, 2005
Hyperbranched Glycopolymers 13
Table 3. Self-Condensing Vinyl Copolymerization of BPEA and AIGlc under Various Conditionsa
solventb
temp (°C)
time (h)
convc (%)
Mn,GPCd (Mw/Mn)
Mn,GPC-viscositye (Mw/Mn)
Rf
DBg
ethyl acetate
ethyl acetate
ethyl acetate
anisole
60
60
80
80
4
24
24
24
48
98
95
58
3100 (1.54)
5180 (1.79)
4800 (1.60)
2900 (1.43)
4800 (1.78)
6600 (1.92)
5300 (2.21)
4400 (1.68)
0.19
0.24
0.23
0.08
0.39
0.42
0.42
0.41
a Copolymerization at a constant comonomer ratio, γ ) [AIGlc]0/[BPEA]0 ) 1.5, and a constant comonomer-to-catalyst ratio, µ ) ([AIGlc]0
+ [BPEA]0)/[CuBr]0 ) 100. b Ethyl acetate or anisole (50 wt % to AIGlc) was used as a solvent. c Conversion of double bonds as determined
by 1H NMR. d Determined by GPC using THF as eluent with PtBuA standards. e Determined by GPC/viscosity measurement. f Mark-
Houwink exponent as determined by GPC/viscosity measurement. g Degree of branching as determined by 1H NMR using eq 2, before
hydrolysis.
( 0.03). As can be seen in Figure 3b, the absolute
intrinsic viscosities of the branched polymers are sig-
nificantly lower than those of the linear one in the
higher molecular weight range (M > 104), suggesting a
more compact architecture with lower intrinsic viscosity.
Note that the viscosity values of the linear polymers are
comparable to those of the branched polymers in the
low molecular weights range (M < 104), suggesting that
the macroscopic quantities such as intrinsic viscosity
and radius of gyration of linear polymers are more or
less similar to those of branched ones due to the bulky
side groups. However, in the higher molecular weight
range, the influence of the bulky side groups is less
significant, and the branched structures leads to com-
pact architectures and a decrease in the viscosity. In
other words, the bulky side group has significant
influence not only on the polymerization rate but also
on the solution behavior in low molecular weight area.
When the same polymerization was carried out in
anisole at 80 °C, the reaction mixture turned brown
after 24 h, and there was an apparent lowering of
molecular weights (Mn,GPC-VISCO ) 4400 and Mw/Mn )
1.68) with lower conversion (58% even after 24 h). The
R value was around 0.08, which is significantly less than
Figure 3. RI signals (a) and Mark-Houwink plots (b) for the
that of typical hyperbranched polymers. Moreover, the
homopolymerization at 80 °C gave a bimodal GPC curve,
as discussed in the previous section. Hence, SCVCP of
AIGlc and BPEA with CuBr/PMDETA in ethyl acetate
using at 60 °C was selected for our further investiga-
tions toward the synthesis of highly branched glyco-
polymers.
polymers obtained by copolymerizations of BPEA and AIGlc
at a constant comonomer ratio, γ ) [AIGlc]0/[BPEA]0 ) 1.5 in
anisole at 80 °C (O), in ethyl acetate (EtOAc) at 80 °C (tilted
3), and in ethyl acetate (EtOAc) at 60 °C (3). The intrinsic
viscosity of linear poly(AIGlc) (s) is given for comparison. See
Table 3 for detailed polymerization conditions.
due to recombination occurring at higher temperatures.
These results suggest that the CuBr/PMDETA system
at 60 °C in ethyl acetate (50 wt % to AIGlc) is a suitable
system for controlled polymerization of AIGlc, but the
increase of the polymerization temperature may lead
to unfavorable side reactions.
Effect of Polymerization Conditions on SCVCP
of BPEA with AIGlc. Copolymerizations were con-
ducted with CuBr/PMDETA in ethyl acetate at a
constant comonomer-to-catalyst ratio, µ ) ([AIGlc]0 +
[BPEA]0)/[CuBr]0 ) 100, and a constant comonomer
ratio, γ ) [AIGlc]0/[BPEA]0 ) 1.5. The results are
summarized in Table 3. The comparisons of the Mark-
Houwink plots are given in Figure 3.
When the SCVCP of AIGlc with BPEA was carried
out in ethyl acetate at 60 °C, 48% conversion was
reached after 4 h. Almost full conversion was reached
only after 24 h, owing to the slow rate of polymerization
of AIGlc as discussed earlier. As can be seen from Table
3, Mn, GPC-VISCO ) 6600 and Mw/Mn ) 1.92 were obtained
at almost full conversion. When the reaction was
conducted in ethyl acetate at 80 °C after 24 h, the
results are comparable to those obtained at 60 °C. In
both cases, the Mark-Houwink exponents of the
branched polymers in THF (R ) 0.23-0.24) are signifi-
cantly lower than that for linear poly(AIGlc) (R ) 0.52
Effect of the Comonomer Ratio on the SCVCP
of BPEA with AIGlc. The effect of the comonomer ratio
on SCVCP of BPEA with AIGlc was investigated with
the CuBr/PMDETA system in ethyl acetate. The copo-
lymerization was carried out at 60 °C at different
comonomer ratios, γ ) [AIGlc]0/[BPEA]0 between 1 and
10, keeping the comonomer-to-catalyst ratio at a con-
stant value of µ ) ([AIGlc]0 + [BPEA]0)/[CuBr]0 ) 100.
In the case of γ ) 1, almost 90% conversion was
achieved after 4 h. However, in the case of γ ) 1.5 the
conversion was 48% even after 4 h. Hence, the polym-
erization time was adjusted, depending on the comono-
mer ratio, γ, because the achievement of almost full
conversion is one of important factors to obtain higher
molecular weights and degree of branching.30 Actually,
the time was prolonged until the reaction mixture
became viscous. For γ values ranging from 2.5 to 10,
almost full conversion was achieved after 24-120 h, as
shown in Table 4. The molecular weights and molecular
weight distribution of the copolymers were characterized
by GPC/viscosity using universal calibration and con-
ventional GPC. In all cases, the molecular weights
determined by GPC/viscosity are higher than the ap-
parent ones obtained by GPC, indicating highly branched
structures.

14 Muthukrishnan et al.
Macromolecules, Vol. 38, No. 1, 2005
Table 4. Self-Condensing Vinyl Copolymerization of BPEA and AIGlc at Different Comonomer Ratios γa
BPEA ratio in polymer
γb
time (h)
convc (%)
Mn,GPCd (Mw/Mn)
Mn,GPC-viscositye (Mw/Mn)
Rf
calcdg
obsd (NMR)h
obsd (EA)i
1
4
24
90
98
91
95
96
7800 (1.43)
5200 (1.79)
4900 (1.49)
5500 (1.48)
5400 (1.37)
9200 (2.84)
6600 (1.92)
5100 (2.13)
11900 (1.59)
13000 (1.95)
0.20
0.24
0.24
0.27
0.28
0.50
0.40
0.28
0.16
0.09
0.41
0.39
0.33
0.19
0.08
0.49
0.46
0.29
0.17
0.10
1.5
2.5
5
24
72
120
10
a Copolymerization at 60 °C with CuBr/PMDETA at a constant comonomer-to-catalyst ratio, µ ) ([AIGlc]0 + [BPEA)0])/[catalyst]0) )
100 in the presence of ethyl acetate (50 wt % to AIGlc). b γ ) [AIGlc]0/[BPEA]0. c Conversion of double bonds as determined by 1H NMR.
d Determined by GPC using THF as eluent with PtBuA standards. e Determined by GPC/viscosity measurement. f Mark-Houwink exponent
as determined by GPC/viscosity measurement. g Calculated from the composition in feed. h Determined by 1H NMR. i Determined from
elemental analysis using the bromine content.
Figure 5. Mark-Houwink plots for the polymers obtained
by copolymerizations of BPEA and AIGlc: γ ) 1 (O), 2.5 (]),
5 (3). The intrinsic viscosity of a linear poly(AIGlc) (s) is given
Figure 4. GPC traces of branched copolymers obtained by
for comparison. Contraction factors, g′ ) [η]branched/[η]linear, for
SCVCP of BPEA and AIGlc at different comonomer ratios, γ
the polymers obtained by SCVCP of BPEA and AIGlc: γ ) 1
) [AIGlc]0/[BPEA]0. See Table 3 for γ ) 1.5 (4 h) and Table 4
(s), 2.5 (- -), 5.0 (‚‚‚). See Table 4 for detailed polymerization
for γ ) 2.5, 5, and 10 for detailed polymerization conditions.
conditions.
As can be seen in Figure 4, the elution curves shift
toward higher molecular weights with increasing comono-
responding polymer, however, is not affected, indicating
that the different fractions have similar branched
structure.
mer ratio, γ. In general, the number-average molecular
weights of branched polymers obtained by SCVCP
increase with γ, and such a tendency was observed in
the SCVCP of tert-butyl acrylate with BPEA38 as well
as in the solution polymerization of 2-(diethylamino)-
ethyl methacrylate with a methacrylate-type inimer.31
Mark-Houwink plots and contraction factors,39 g′ )
[η]branched/[η]linear, as a function of the molecular weight
for representative branched polymers obtained by SCVCP
are shown in Figure 5. Relationships between dilute
solution viscosity and molecular weight have been
determined, and the Mark-Houwink constant typically
varies between 0.28 and 0.2, depending on the degree
of branching. In contrast, the exponent is typically in
the region of 0.6-0.8 for linear homopolymers in a good
solvent with a random coil conformation. The Mark-
Houwink exponent of the mixture of linear poly(AIGlc)s
(R ) 0.52 ( 0.03) is lower than that of poly(tert-butyl
acrylate) (R ) 0.80), indicating less favorable interaction
with the solvent. In both cases, the exponent values are
well within the range of linear homopolymers. The
contraction factor is another way of expressing the
compact structure of a branched polymer. From Figure
5, it is obvious that the viscosities of branched poly-
(AIGlc)s are significantly lower than those of the linear
one and increase with γ. There were only slight differ-
ences in the slopes between the branched copolymers,
but the values at each molecular weight increase with
increasing γ value. The contraction factors for all the
branched copolymers decrease with increasing molec-
ular weights. These observations indicate that the
differences in the molecular weights obtained from the
GPC/viscosity compared to conventional GPC arises
from a systematic decrease in Mark-Houwink expo-
nent, R, and the contraction factor, g′, due to a compact
In this study, molecular weights up to Mn,GPC-VISCO
)
13 000 could be obtained at γ ) 10, whereas the
polymers obtained at lower γ values (γ ) 1.5, 2.5) had
lower molecular weights, which are in accordance with
the general tendency. A slight deviation from the
tendency (for example, when γ ) 1.5, 2.5, and 5,
Mn,GPC-VISCO ) 6600, 5100, and 11 900, respectively)
may be related to the difference in the conversion
observed at each γ value. According to theory29 (assum-
ing equal reactivity of active centers), the number-
average degree of polymerization increases drastically
with conversion of the comonomer, especially at the end
of the polymerization. We believe that the relationship
between the molecular weights and γ is independent of
the polymerization and catalyst systems and may be
attributed to an inherent tendency for the cyclization
process in SCVCP. The SCVCP at γ > 10 is considered
to be possible, but a prolonged reaction time or higher
catalyst concentration is required to attain higher
conversion in the cases of higher γ. This is an inherent
character of the system used in this study, which is
basically due to the nature of the bulky sugar-carrying
acrylate. It must be noted that multimodal GPC traces
obtained with γ ) 5 and 10 (Figure 4) indicate the
presence of fractions with different hydrodynamic vol-
umes. The Mark-Houwink plot (Figure 5) of the cor-

Macromolecules, Vol. 38, No. 1, 2005
Hyperbranched Glycopolymers 15
Figure 6. Dependence of the Mark-Houwink exponent, R,
on comonomer ratio, γ (9). (b) Linear poly(AIGlc).
Figure 8. 1H NMR spectra (CDCl3) of the linear (a) and
branched poly(AIGlc): γ ) 5 (b), 2.5 (c), 1 (d), and poly(BPEA)
(e).38
Figure 7. Dependence of bromine contents on 1/(γ + 1).
Samples: see Table 4. The calculated value (s) is given for
comparison.
poly(AIGlc). In the case of the copolymers, besides the
signal of poly(AIGlc) segment, the BPEA inimer signals
appear at 4.0-4.5 ppm attributed to the protons of the
ethylene linkage and the protons which are geminal to
bromine in either A*, B*, or M*, all which are derived
from BPEA. The latter protons correspond to the end
groups. Apart from the peak, a large doublet at 1.85
ppm in the copolymers is assigned to CH3 of the
2-bromopropionyloxy group, B*, while the peak around
1.1-1.3 ppm is assigned to b, which is formed by
addition of the monomer to B*. This peak b is overlap-
ping with the isopropylidene protons of the poly(AIGlc)
segment. The BPEA content in the copolymers obtained
structure resulting polymer from the increased number
of branches.
The effect of the comonomer ratio, γ, on the Mark-
Houwink exponent is shown in Figure 6. In the whole
range of γ values, the exponents of the branched
polymers are significantly lower (R ) 0.20-0.28) com-
pared to that for linear poly(AIGlc) (R ) 0.52 ( 0.03).
Such a decrease in the Mark-Houwink exponent, R, as
well as the contraction factor, g′, provides conclusive
evidence for a more compact architecture with increas-
ing number of branches and confirms our earlier obser-
vations in SCVCP of other (meth)acrylates.30,31
1
by SCVCP was determined from the H NMR spectra
With ATRP, nearly every chain should contain a
halogen atom at its end group, if termination and
transfer are essentially absent. In other words, the
existence of a halogen atom at the chain end can be used
as an indicator to confirm that the polymerization
proceeds in controlled fashion without termination and
transfer reactions. The halogen atom can be also
replaced through a variety of reactions leading to end-
functional polymers and used as the initiating part for
polymerization of a second monomer. In the case of ideal
SCVCP via ATRP, the resulting branched polymers
carry one bromoester function per inimer unit, and the
functionality decreases with comonomer (AIGlc) com-
position. As shown in Figure 7, the bromine contents of
the branched polymers are dependent upon the comono-
mer composition in the feed and are in fair agreement
with the calculated values. This result indicates the
existence of reasonable number of bromine atoms at the
chain ends, suggesting that unfavorable termination
and transfer reactions are essentially negligible under
the condition used in this study, and the number of
bromoester end groups can be simply determined by the
comonomer ratio, γ, in the feed.
by comparing the peaks at 3.9-4.6 ppm attributed to
the sum of five protons (A, B, C, G) of the poly(AIGlc)
segment and five protons of the ethylene linkage and
that geminal to bromine as mentioned above and the
peak at 5.0-5.3 ppm corresponding to one proton (F) of
the poly(AIGlc) segment, as shown in Figure 8. Thus,
the comonomer composition can be calculated using
eq 1
5H(x) + 5H(1 - x)
1H(x)
integral at 3.9-4.6 ppm
integral at 5.0-5.3 ppm
)
(1)
where “x” is the fraction of the monomer, whereas “1 -
x” is the fraction of the inimer in the polymer. The
comonomer fractions calculated from the ratio of these
peaks are in good agreement with the comonomer
composition in the feed, which corresponds to the γ
value, as can be seen in Table 4. The comonomer
fractions were also determined by elemental analysis
via bromine content, which are also consistent with the
theoretical values within experimental error. The agree-
ments suggest complete inimer incorporation.
The structure of the linear and branched poly(AIGlc)s
Degree of Branching (DB). Figure 8 shows 1H
NMR spectra with the complete assignment of the linear
and branched poly(AIGlc)s as well as poly(BPEA) ob-
tained by a homo-SCVP of BPEA. For poly(BPEA), the
proportions of b and B* can be calculated from the broad
peak at 1.0-1.3 ppm assigned to b and large doublet at
was also confirmed by 1H NMR and elemental analysis.
1
Figure 8 shows the respective H NMR spectra of the
poly(AIGlc)s obtained by ATRP and SCVCP. The char-
acteristic peaks at 1.2-1.4 (isopropylidene protons),
3.9-4.6, and 5.0-5.3 ppm are clearly seen in the linear

16 Muthukrishnan et al.
Macromolecules, Vol. 38, No. 1, 2005
Figure 9. 1H NMR spectra (DMSO-d6) of the polymers obtained after deprotection of linear poly(AIGlc) (a) and branched poly-
(AIGlc)s: γ ) 2.5 (b), 1.5 (c).
Table 5. Degree of Branching of Copolymers Obtained by
1.85 ppm assigned to B*. In the cases of the copolymers
SCVCP of AIGlc and BPEA at Different Comonomer
obtained by SCVCP, these peaks should be related to
Ratios γ
the degree of branching and the comonomer composi-
f
γa
bb
bc
DBd
DBe
DBtheo
tion. In our copolymerization system, since the protons
of the isopropylidene groups overlap with the b protons,
we have indirectly calculated the proportion of b using
1H NMR. For equal reactivity of active sites, the degree
of branching determined by NMR, DBNMR, at full
conversion is given as29
1.0
1.5
2.5
0.65
0.75
0.78
0.60
0.67
0.68
0.43
0.42
0.35
0.42
0.38
0.31
0.49
0.47
0.40
a γ ) [AIGlc]0/[BPEA]0. b Fraction of reacted B* units as deter-
mined by 1H NMR before hydrolysis. c Fraction of reacted B* units
as determined by 1H NMR after hydrolysis. d Degree of branching
as determined by 1H NMR using eq 2 before hydrolysis. e Degree
of branching as determined by 1H NMR using eq 2 after hydrolysis.
f Theoretical degree of branching as determined using eq 3.
b
b
DBNMR ) 2
1 -
(2)
(
)
[
(
)
]
γ + 1
γ + 1
According to the theory of SCVCP, DBtheo, at full
conversion, can be represented as
agreement with the values of b obtained by the indirect
method, as can be seen in Table 5.
DBNMR decreases with γ, as predicated by calcula-
tions. In all the cases, however, the observed values are
slightly lower than the calculated ones, which might be
attributed to the simplifications made for the calcula-
tions, i.e., equal reacativity of A*, B*, and M* chain
ends. Although NMR experiments afford a conclusive
measurement of the degree of branching for lower γ
values, the low concentration of branch points in the
copolymer at γ > 5 does not allow the determination of
the degree of branching directly by the spectroscopic
method because of low intensities of the peaks in the
regions around 1.85 ppm and 1.0-1.3 ppm, as shown
in Figure 9. However, for the case of high comonomer
ratios, γ . 1, the relation between DBtheo and γ becomes
very simple and does not depend on the reactivity ratios
of the various active centers and is represented as DBtheo
≈ 2/(γ + 1).29 When the reactivities of the various active
centers are not equal, the dependences are more com-
plex, and the degree of branching may be higher or
lower, depending upon the systems. In this copolymer-
ization system, the rate constant of the sugar-carrying
acrylate was comparatively lower than that of the
normal acrylates, and longer polymerization time was
required to attain almost full conversion. Hence, the rate
constant of BPEA used as an AB* inimer should be
different from that of the comonomer, AIGlc, used in
this study. This could also lead to the formation of
starlike structures instead of randomly branched ar-
2(1 - e-(γ+1))(γ + e-(γ+1)
(γ + 1)2
)
DBtheo
)
(3)
The fraction of b units could be calculated by comparing
the peaks at 3.6-4.5 ppm and the large doublet at 1.85
ppm in the copolymers ranging from γ ) 1 to 2.5. The
peaks at 3.6-4.5 ppm correspond to five protons of poly-
(AIGlc) segment and five protons of the inimer, as
mentioned earlier. The doublet at 1.85 ppm corresponds
to three protons of the methyl group of B* (2-bromopro-
pionyloxy group). Taking into account the inimer (BPEA)
composition in the copolymer (see eq 1), we can calculate
the integral of the peak at 1.85 ppm, corresponding to
B*. On the other hand, the actual ratio of the peaks at
1
1.85 and 3.6-4.5 ppm determined by H NMR provide
the value of B* - b. From these approaches, DBNMR
)
0.43 and DBtheo ) 0.49 can be obtained at γ ) 1 (b )
0.65). The reliability of the method for the evalua-
tion of b was verified by comparing the value of
b after the hydrolysis of the branched and homo poly-
(AIGlc)s. After the hydrolysis, the isopropylidene groups
disappear, and the broad peak around 1.0-1.3 ppm is
clearly visible (Figure 9), which is then compared with
the B* protons for the evaluation of the proportion of b
using the equation b ) (region at 1.1-1.3 ppm)/(sum of
region at 1.85 ppm (B*) and region at 1.1-1.3 ppm (b)).
The values of b obtained after the hydrolysis are in good

Macromolecules, Vol. 38, No. 1, 2005
Hyperbranched Glycopolymers 17
weights and the branched architecture of the hydrolyzed
products correspond to those of the branched poly-
(AIGlc)s. The molecular weight and molecular weight
distribution of the hydrolyzed homopolymer is Mn
)
22 900 (Mw/Mn ) 1.16), which is fairly in agreement
with Mn,calcd ) 23 400 and also with the results Mn )
24 200 (Mw/Mn ) 1.14) obtained before hydrolysis.
FT-IR spectra of the linear and branched poly(AGlc)s
obtained after hydrolysis are shown in Figure S-2 (see
Supporting Information). The absorption bands due to
the isoproylidene groups are observed at 3000, 1450, and
1380 cm-1 assigned to the C-H stretching and C-H
asymmetric and symmetric deformation modes, respec-
tively. The C-O stretching band of the isopropylidene
group is also observed in the region between 1200 and
1000 cm-1. After the deprotection, these bands disap-
pear, and a strong absorption band around 3500 cm-1
is observed due to the hydroxyl group formed by the
deprotection. The spectrum of the branched poly(AGlc)
at γ ) 2.5 having less branch points is almost same as
that of the linear poly(AGlc).
Figure 10. NMP phase GPC traces of (a) linear poly(AGlc)
and (b) branched poly(AGlc)s.
chitectures. A detailed investigation is now in progress
and will be reported elsewhere.
Conclusions
Self-condensing vinyl copolymerization (SCVCP) of
the sugar-carrying acrylate, AIGlc, with the acrylic AB*
inimer via ATRP provides a straightforward strategy
for generating water-soluble hyperbranched glycopoly-
mers and their precursors. Linear poly(AIGlc) with
controlled molecular weights and very low polydispersi-
ties (1.05 < PDI < 1.14) were obtained in a controlled
fashion using the CuBr/PMDETA catalyst system. The
significant influence of the bulky isopropylidene-pro-
tected sugar moiety in AIGlc was observed on the
polymerization rate and the macroscopic quantities such
as intrinsic viscosity and radius of gyration. It was found
that a suitable choice of the polymerization conditions
is required to obtain water-soluble glycopolymers with
characteristic highly branched architectures, which can
then be manipulated for various biological, pharmaceu-
tical, and medical applications. Further studies on this
interesting branched glycopolymers for such directions
will be reported separately. This work broadens the
scope of facile synthesis of branched glycopolymers by
a controlled polymerization technique, in which the
architecture can be simply manipulated by the nature
and composition of sugar-carrying vinyl monomer used
as a comonomer. Extension to planar and spherical
polymer brushes is also planned.
Deprotection of Linear and Branched Poly-
(AIGlc)s. The hydrolysis of the isopropylidene groups
in the linear and branched poly(AIGlc)s was performed
by treating the samples with formic acid.22 The final
product was obtained by freeze-drying from dioxane
after the deprotected polymer was dialyzed against
water. Figure 9a reveals that the signals of the isopro-
pylidene protons (1.2-1.4 ppm) completely disappear
after the hydrolysis of the linear poly(AIGlc), and a
broad signal assignable to anomeric hydroxyl groups of
the sugar moieties (6.4-7.0 ppm) appears. This proves
that the deprotection proceeds quantitatively. The linear
deprotected polymers, poly(3-O-acryloyl-R,â-D-glucopy-
ranoside), which can be abbreviated as poly(AGlc), are
white powders completely soluble in water, methanol,
and DMSO but insoluble in THF and acetone. For the
branched poly(AGlc)s, the solubility of the polymers
were dependent upon the comonomer ratios, γ. In the
case of γ ) 1, the branched poly(AGlc) was partially
soluble in water, which is attributed to the 50% of the
nonpolar inimer segment. From γ ) 1.5 onward, the
polymers were completely soluble in water, DMSO, and
methanol but insoluble in THF and acetone. It is
important to note that the unchanged resonance signal
of protons of the ethylene linkage and those geminal to
bromine at 4.0-4.6 ppm suggests that the BPEA
composition in the branched poly(AGlc) is almost the
same as that in branched poly(AIGlc). As mentioned in
the previous section, the fraction of b units (reacted B*
units) and DBNMR after the hydrolysis are in good
agreement with those of the branched poly(AIGlc)s.
These results also indicate that the branched structure
is intact during the complete deprotection of the iso-
propylidene groups and proceeds selectively to yield
characteristic branched poly(AGlc)s.
Acknowledgment. We appreciate S. Wunder for her
help with the GPC-viscosity measurements.
Supporting Information Available: Figures S-1 and S-2.
This material is available free of charge via the Internet at
http://pubs.acs.org.
References and Notes
(1) Lee, Y. C.; Lee, R. T. Acc. Chem. Res. 1995, 28, 321-327.
(2) Bovin, N. V.; Gabius, H. J. Chem Soc. Rev. 1995, 24, 413.
(3) Capila, I.; Linhardt, R. J. Angew. Chem., Int. Ed. 2002, 41,
391-412.
Figure 10 represents the NMP-phase GPC traces of
the linear and branched poly(AGlc)s. The hyper-
branched samples show a broad distribution and com-
parable molecular weights as observed in THF-phase
GPC before the hydrolysis. The molecular weights and
molecular weight distributions for the hydrolyzed
branched poly(AGlc)s at γ ) 1.0 and 1.5 are Mn ) 7900
(Mw/Mn ) 1.91) and Mn ) 7300 (Mw/Mn ) 1.40),
respectively. These results suggest that the molecular
(4) Klein, J.; Kunz, M.; Kowalczyk, J. Makromol. Chem. 1990,
191, 517-528.
(5) Gunay, N. S.; Linhardt, R. J. Planta Med. 1999, 65, 301-
306.
(6) Simanek, E. E.; McGarvey, G. J.; Jablonowski, J. A.; Wong,
C. H. Chem. Rev. 1998, 98, 833-862.
(7) Petronio, M. G.; Mansi, A.; Gallinelli, C.; Pisani, S.; Seganti,
L.; Chiarini, F. Chemotherapy 1997, 43, 211-217.

18 Muthukrishnan et al.
Macromolecules, Vol. 38, No. 1, 2005
(8) Ladmiral, V.; Melia, E.; Haddleton, D. M. Eur. Polym. J.
2004, 40, 431-449.
(24) Liang, Y. Z.; Li, Z. C.; Chen, G. Q.; Li, F. M. Polym. Int. 1999,
48, 739-742.
(25) Li, Z. C.; Liang, Y. Z.; Chen, G. Q.; Li, F. M. Macromol. Rapid
Commun. 2000, 21, 375-380.
(26) Ejaz, M.; Ohno, K.; Tsujii, Y.; Fukuda, T. Macromolecules
2000, 33, 2870-2874.
(9) Wang, Q.; Dordick, J. S.; Linhardt, R. J. Chem. Mater. 2002,
14, 3232-3244.
(10) Dordick, J. S.; Linhardt, R. J.; Rethwisch, D. G. CHEMTECH
1994, 24, 33-39.
(11) Chen, X. M.; Dordick, J. S.; Rethwisch, D. G. Macromolecules
1995, 28, 6014-6019.
(12) Zanini, D.; Roy, R. J. Org. Chem. 1998, 63, 3486-3491.
(13) Dimick, S. M.; Powell, S. C.; McMahon, S. A.; Moothoo, D.
N.; Naismith, J. H.; Toone, E. J. J. Am. Chem. Soc. 1999,
121, 10286-10296.
(14) Roy, R.; Pon, R. A.; Tropper, F. D.; Andersson, F. O. J. Chem.
Soc. Chem. Comm. 1993, 264-265.
(15) Aoi, K.; Itoh, K.; Okada, M. Macromolecules 1995, 28, 5391-
5393.
(27) Narain, R.; Armes, S. P. Chem. Commun. 2002, 2776-2777.
(28) Narain, R.; Armes, S. P. Macromolecules 2003, 36, 4675-
4678.
(29) Litvinenko, G. I.; Simon, P. F. W.; Mu¨ller, A. H. E. Macro-
molecules 1999, 32, 2410-2419.
(30) Simon, P. F. W.; Mu¨ller, A. H. E. Macromolecules 2001, 34,
6206-6213.
(31) Mori, H.; Walther, A.; Andre, X.; Lanzendorfer, M. G.; Mu¨ller,
A. H. E. Macromolecules 2004, 37, 2054-2066.
(32) Litvinenko, G. I.; Simon, P. F. W.; Muller, A. H. E. Macro-
molecules 2001, 34, 2418-2426.
(33) Mori, H.; Mu¨ller, A. H. E. In Dendrimers V, Topics in Current
Chemistry; Vo¨gtle, F., Ed.; Springer: Heidelberg, 2003; Vol.
228, pp 1-37.
(34) Matyjaszewski, K.; Gaynor, S. G.; Kulfan, A.; Podwika, M.
Macromolecules 1997, 30, 5192-5194.
(35) Mori, H.; Bo¨ker, A.; Krausch, G.; Mu¨ller, A. H. E. Macromol-
ecules 2001, 34, 6871-6882.
(16) Yamada, K.; Yamaoka, K.; Minoda, M.; Miyamoto, T. J.
Polym. Sci., Polym. Chem. 1997, 35, 255-261.
(17) Aoi, K.; Tsutsumiuchi, K.; Okada, M. Macromolecules 1994,
27, 875-877.
(18) Fraser, C.; Grubbs, R. H. Macromolecules 1995, 28, 7248-
7255.
(19) Wulff, G.; Schmid, J.; Venhoff, T. Macromol. Chem. Phys.
1996, 197, 259-274.
(20) Ohno, K.; Tsujii, Y.; Miyamoto, T.; Fukuda, T.; Goto, M.;
Kobayashi, K.; Akaike, T. Macromolecules 1998, 31, 1064-
1069.
(36) Ouchi, T.; Jokei, S.; Chikashita, H. J. Heterocycl. Chem. 1982,
19, 935-936.
(37) Benoˆıt, H.; Grubisic, Z.; Rempp, P.; Decker, D.; Zilliox, J. G.
J. Chem. Phys. 1966, 63, 1507.
(21) Ohno, K.; Tsujii, Y.; Fukuda, T. J. Polym. Sci., Polym. Chem.
1998, 36, 2473-2481.
(22) Ohno, K.; Izu, Y.; Yamamoto, S.; Miyamoto, T.; Fukuda, T.
Macromol. Chem. Phys. 1999, 200, 1619-1625.
(23) Gotz, H.; Harth, E.; Schiller, S. M.; Frank, C. W.; Knoll, W.;
Hawker, C. J. J. Polym. Sci., Polym. Chem. 2002, 40, 3379-
3391.
(38) Mori, H.; Chan Seng, D.; Lechner, H.; Zhang, M.; Mu¨ller, A.
H. E. Macromolecules 2002, 35, 9270-9281.
(39) Burchard, W. Adv. Polym. Sci. 1999, 143, 113-194.
MA048436C