Mechanistic insights for block copolymer morphologies: How do worms form vesicles?

Article abstract of DOI:10.1021/ja206301a

Amphiphilic diblock copolymers composed of two covalently linked, chemically distinct chains can be considered to be biological mimics of cell membrane-forming lipid molecules, but with typically more than an order of magnitude increase in molecular weight. These macromolecular amphiphiles are known to form a wide range of nanostructures (spheres, worms, vesicles, etc.) in solvents that are selective for one of the blocks. However, such self-assembly is usually limited to dilute copolymer solutions (<1%), which is a significant disadvantage for potential commercial applications such as drug delivery and coatings. In principle, this problem can be circumvented by polymerization-induced block copolymer self-assembly. Here we detail the synthesis and subsequent in situ self-assembly of amphiphilic AB diblock copolymers in a one pot concentrated aqueous dispersion polymerization formulation. We show that spherical micelles, wormlike micelles, and vesicles can be predictably and efficiently obtained (within 2 h of polymerization, >99% monomer conversion) at relatively high solids in purely aqueous solution. Furthermore, careful monitoring of the in situ polymerization by transmission electron microscopy reveals various novel intermediate structures (including branched worms, partially coalesced worms, nascent bilayers, "octopi", "jellyfish", and finally pure vesicles) that provide important mechanistic insights regarding the evolution of the particle morphology during the sphere-to-worm and worm-to-vesicle transitions. This environmentally benign approach (which involves no toxic solvents, is conducted at relatively high solids, and requires no additional processing) is readily amenable to industrial scale-up, since it is based on commercially available starting materials.

Full text of DOI:10.1021/ja206301a

ARTICLE
pubs.acs.org/JACS
Mechanistic Insights for Block Copolymer Morphologies:
How Do Worms Form Vesicles?
Adam Blanazs,† Jeppe Madsen,† Giuseppe Battaglia,‡ Anthony J. Ryan,† and Steven P. Armes*,†
†Department of Chemistry, The University of Sheffield, Brook Hill, Sheffield S3 7HF, United Kingdom
‡Department of Biomedical Science, The University of Sheffield, Western Bank, Sheffield S10 2TN, United Kingdom
S Supporting Information
b
ABSTRACT: Amphiphilic diblock copolymers composed of
two covalently linked, chemically distinct chains can be con-
sidered to be biological mimics of cell membrane-forming lipid
molecules, but with typically more than an order of magnitude
increase in molecular weight. These macromolecular amphi-
philes are known to form a wide range of nanostructures
(spheres, worms, vesicles, etc.) in solvents that are selective
for one of the blocks. However, such self-assembly is usually
limited to dilute copolymer solutions (<1%), which is a signif-
icant disadvantage for potential commercial applications such as
drug delivery and coatings. In principle, this problem can be circumvented by polymerization-induced block copolymer self-
assembly. Here we detail the synthesis and subsequent in situ self-assembly of amphiphilic AB diblock copolymers in a one pot
concentrated aqueous dispersion polymerization formulation. We show that spherical micelles, wormlike micelles, and vesicles can
be predictably and efficiently obtained (within 2 h of polymerization, >99% monomer conversion) at relatively high solids in purely
aqueous solution. Furthermore, careful monitoring of the in situ polymerization by transmission electron microscopy reveals various
novel intermediate structures (including branched worms, partially coalesced worms, nascent bilayers, “octopi”, “jellyfish”, and
finally pure vesicles) that provide important mechanistic insights regarding the evolution of the particle morphology during the
sphere-to-worm and worm-to-vesicle transitions. This environmentally benign approach (which involves no toxic solvents, is
conducted at relatively high solids, and requires no additional processing) is readily amenable to industrial scale-up, since it is based
on commercially available starting materials.
’ INTRODUCTION
copolymer vesicles that can respond to external stimuli such as
pH,22,23 oxidation,24 and temperature25 offer considerable po-
tential for intracellular drug delivery.16,26 However, in many cases,
the processing route (e.g., solvent exchange, film rehydration,
pH switch, etc.) used to produce block copolymer aggregates
can influence their final morphology to a similar extent as the
hydrophobic/hydrophilic balance of the copolymer.27 Further-
more, such processing techniques are normally limited to dilute
copolymer solutions (<1%), which is a significant disadvantage
for commercial viability.
The aqueous self-assembly of amphiphilic lipids to form cell
membranes is essential for the existence of life, as it is a prereq-
uisite for cellular compartmentalization, gene segregation, and
the generation of chemical gradients.1 Small molecule surfactant
self-assembly in aqueous solution has been extensively studied
for many decades.2,3 It is well-known that the surfactant concen-
tration and itshydrophilic/hydrophobic balancedictatethe precise
nature of the various nanostructures formed in aqueous solution
(i.e., spherical micelles, wormlike micelles, or vesicles).2ꢀ4
Polymerization-induced self-assembly is well established in
the bulk for toughened thermosets and elastomers formed by
step polymerization.28,29 However, as far as we are aware, there
are very few reports30ꢀ32 of polymerization-induced self-assem-
bly by living radical polymerization under aqueous dispersion poly-
merization conditions. This is not particularly surprising: most
water-miscible monomers do not form water-insoluble polymers,
which is an essential prerequisite for such formulations. Recently,
we reported the reversible additionꢀfragmentation chain transfer
Controlled polymerization techniques (such as ionic and “living”
radical polymerization) have enabled the synthesis of a wide
range of amphiphilic block copolymers that can self-assemble to
form spherical micelles,5 wormlike micelles,6ꢀ9 toroids,10 micel-
lar networks,11 and vesicles (aka polymersomes)5,12ꢀ15 for various
applications.16 Compared to surfactant micelles, block copoly-
mer aggregates are much more robust: the exchange kinetics
of individual chains between aggregates is typically significantly
slower, leading in many cases to nonergodic (i.e., kinetically
frozen) systems.17,18 Moreover, higher molecular weights and
chain entanglements lead to thicker, more resilient vesicular
membranes compared to lipid counterparts.19ꢀ21 In particular,
Received: July 7, 2011
Published: August 17, 2011
r
2011 American Chemical Society
16581
dx.doi.org/10.1021/ja206301a J. Am. Chem. Soc. 2011, 133, 16581–16587
|

Journal of the American Chemical Society
ARTICLE
(RAFT) aqueous dispersion polymerization of 2-hydroxypropyl
methacrylate (HPMA) using a water-soluble poly(glycerol mono-
methacrylate) (PGMA) chaintransferagent(CTA).33 Importantly
for this formulation, HPMA monomer is water-miscible up to
13 w/v % at room temperature, yet forms a water-insoluble
polymer. In the present work, we further investigate this for-
mulation. We show that spherical micelles, wormlike micelles,
and vesicles are predictably obtained for HPMA polymerizations
conducted at 10 w/v %. Complete monomer conversions are
achieved at 70 °C within 2 h, and the particles comprise near-
monodisperse copolymer chains. Furthermore, careful monitor-
ing of the in situ polymerization by transmission electron
microscopy (TEM) reveals various intermediate structures
(including branched worms, partially coalesced worms, nascent
bilayers, “octopi”, “jellyfish”, and finally pure vesicles) that pro-
vide important mechanistic insights regarding the evolution of
the particle morphology.
monomer (0.4462 g, 3.1 mmol, Aldrich) were weighed into a 25 mL round-
bottomed flask and purged with N2 for 20 min. ACVA was added
(1.8 mg, 0.0063 mmol, CTA/ACVA molar ratio = 3:1) and purged with
N2 for a further 5 min. Deionized water (5.4 mL, 10 w/v %), which had
been purged with N2 for 30 min, was then added, and the solution was
degassed for a further 5 min prior to immersion in an oil bath set at 70 °C.
The reaction solution was stirred overnight (16 h) to ensure complete
HPMA monomer conversion and quenched with exposure to air.
Kinetics of the Aqueous Dispersion Polymerization of
HPMA for the Sphere-to-Worm-to-Vesicle Transitions When
Targeting PGMA47-PHPMA200. PGMA47 macro-CTA (0.200 g, 0.026
mmol), HPMA monomer (0.7437 g, 5.2 mmol, Aldrich), and sodium 2,2-
dimethyl-2-silapentane-5-sulfonate (DSS) (0.0225 g, 0.10 mmol, HPMA/
DSS molar ratio = 50:1) were weighed into a 25 mL round-bottomed
flask and purged with N2 for 20 min. ACVA was added (1.2 mg, 0.004
mmol, CTA/ACVA molar ratio = 6:1) and purged with N2 for a further
5 min. Deionized water (8.5 mL, 10 w/v %), which had been purged with
N2 for 30 min, was then added, a sample was immediately taken for 1H
NMR analysis, and the solution was degassed for a further 5 min prior to
immersion in an oil bath set at 70 °C. The “zero time” (t = 0 min) for this
polymerization was arbitrarily taken to be the point when the degassed
reaction solution was first immersed in an oil bath set at 70 °C, rather
than the time taken for the reaction solution to attain this temperature.
Aliquots were then removed via syringe at various time intervals for 1H
NMR and TEM analysis. 1H NMR samples were quenched by dilution
in D2O at 20 °C. Monomer conversions were normalized using the DSS
as an internal standard, and are expressed relative to the ratio of monomer
to DSS observed at “zero time”. For TEM analysis, aliquots were diluted
50-fold with water at 20 °C to generate 0.20 wt % dispersions.
’ MATERIALS AND METHODS
Materials. Glycerol monomethacrylate (GMA; 99.8%) was donated
by Cognis Performance Chemicals (Hythe, U.K.) and used without
further purification. 2-Hydroxypropyl methacrylate (HPMA) was do-
nated by Cognis Performance Chemicals and was also purchased from
Sigma Aldrich; in each case, monomer was passed through a DHR-4
inhibitor removal column (Scientific Polymer Products, Ontario, NY)
prior to use. 2-Cyano-2-propyl dithiobenzoate (CPDB), 4,40-azobis(4-
cyanopentanoic acid) (ACVA; V-501; 99%) D2O, anhydrous ethanol
(99%), methacrylic anhydride (94%), N,N-dimethylaminopyridine
(99%), and dialysis tubing (1 kD molecular weight cutoff) were pur-
chased from Sigma Aldrich U.K. and were used as received. In the case of
Polymer Characterization. 1H NMR Spectroscopy. All NMR
spectra were recorded on a 400 MHz Bruker Avance-400 spectrometer
(64 scans averaged per spectrum).
1
the CPDB, the manufacturer’s stated purity was 97%, but H NMR
Gel Permeation Chromatography (GPC). Copolymer molecular
weights and polydispersities were determined using a DMF GPC setup
operating at 60 °C and comprising two Polymer Laboratories PL gel
5 μm Mixed C columns connected in series to a Varian 390 LC
multidetector suite (refractive index detector) and a Varian 290 LC
pump injection module. The GPC eluent was HPLC grade DMF
containing 10 mM LiBr at a flow rate of 1.0 mL minꢀ1. Dimethyl
sulfoxide (DMSO) was used as a flow-rate marker. Calibration was
conducted using a series of 10 near-monodisperse poly(methyl
methacrylate) standards (Mn = 625ꢀ618 000 g molꢀ1). The chromato-
grams were analyzed using Varian Cirrus GPC software (version 3.3).
Dynamic Light Scattering. Intensity-average hydrodynamic dia-
meters of the dispersions were obtained by DLS using a Malvern
Zetasizer NanoZS instrument. Aqueous dispersions of 0.20 w/v % were
analyzed using disposable cuvettes, and all data were averaged over three
consecutive runs.
Transmission Electron Microscopy (TEM). Aggregate solutions were
diluted 50-fold at 20 °C to generate 0.20 w/v % dispersions. Copper/
palladium TEM grids (Agar Scientific) were surface-coated in-house to
yield a thin film of amorphous carbon. The grids were then plasma glow-
discharged for 30 s to create a hydrophilic surface. Individual samples
(0.20 w/v %, 12 μL) were adsorbed onto the freshly glow-discharged
grids for 1 min and then blotted with filter paper to remove excess
solution. To stain the aggregates, uranyl formate (0.75 w/v %) solution
(9 μL) was soaked on the sample-loaded grid for 20 s and then carefully
blotted to remove excess stain. The grids were then dried using a vacuum
hose. Imaging was performed on a Phillips CM100 instrument at 100 kV,
equipped with a Gatan 1 k CCD camera.
analysis indicated a purity of only 75%. This reduced purity was taken in
account when calculating the target degree of polymerization for the
PGMA block. The water-soluble internal standard used in the NMR
experiments, sodium 2,2 dimethyl-2-silapentane-5-sulfonate (DSS),
and CD3OD were purchased from Goss Scientific (Nantwich, U.K.).
Triethylamine, magnesium sulfate, sodium hydrogen carbonate, and
sodium chloride were of Laboratory Reagent grade and purchased from
Fisher Scientific (Loughborough, U.K.). All solvents were of HPLC
quality and purchased from Fisher Scientific (Loughborough, U.K.).
Synthesis of the PGMA47 Macro-CTA. CPDB RAFT agent (1.5
mmol, 0.33 g, purchased from Sigma Aldrich with 75% purity as judged
by 1H NMR spectroscopy) and GMA monomer (89.6 mmol, 14.35 g)
were weighed into a 50 mL round-bottomed flask and purged under N2
for 20 min. ACVA (0.30 mmol, 83.7 mg, CTA/ACVA molar ratio = 5:1)
and anhydrous ethanol (40 w/v %), which had been purged with N2 for
30 min, were then added, and the resulting red solution was purged for a
further 10 min. The sealed flask was immersed into an oil bath set at
70 °C for 80 min (GMA conversion 57%, see Supporting Information
Figure S1) and quenched in liquid nitrogen. Methanol (50 mL) was
added to the reaction solution, followed by precipitation into a 10-fold
excess of cyclohexane (1 L). The precipitated PGMA macro-CTA was
washed three times with cyclohexane and then dialyzed against metha-
nol overnight (with three changes of methanol) using semipermeable
cellulose tubing (SPECTRA/POR, corresponding to a molecular weight
cutoff of 1000). 1H NMR indicated a degree of polymerization of 47 for
the PGMA macro-CTA. Mn = 14,100 and Mw/Mn = 1.13, as judged by
GPC using DMF eluent, a refractive index detector, and a series of near-
monodisperse poly(methyl methacrylate) calibration standards.
Reverse-Phase High-Performance Liquid Chromatography (HPLC).
HPLC was utilized to quantify the relative amounts of a dimethacrylate
impurity within the HPMA monomer. The experimental setup consisted
of an autosampler (Varian model 410), a solvent delivery module
RAFT Aqueous Dispersion Polymerization of PGMA47-
PHPMA160. A typical protocol for the synthesis of PGMA47-PHPMA160
is as follows: PGMA47 macro-CTA (0.150 g, 0.019 mmol) and HPMA
16582
dx.doi.org/10.1021/ja206301a |J. Am. Chem. Soc. 2011, 133, 16581–16587

Journal of the American Chemical Society
ARTICLE
Figure 2. DMF gel permeation chromatograms (vs poly(methyl
methacrylate) standards) obtained for G112-H1000 diblock copolymers
and the corresponding G112 macro-CTA synthesized via RAFT aqueous
dispersion polymerization at 70 °C and 10 w/v % solids. The red trace
illustrates the influence of small amounts of a dimethacrylate impurity
within the HPMA monomer, resulting in light branching and a relatively
high polydispersity. However, removal of this dimethacrylate impurity
from the HPMA monomer via column chromatography prior to
copolymer synthesis leads to a near-monodisperse G112-H1000 diblock
copolymer (blue trace).
controlled synthesis of relatively high molecular weight diblock
copolymers. For example, G112-H1000 was readily synthesized
with more than 99% HPMA conversion again being obtained
within 2 h at 70 °C (see Figure 2 and Supporting Information
Figure S4). Good blocking efficiencies were observed, but a rather
Figure 1. (a) Chemical structure of the poly(glycerol monometh-
acrylate)ꢀpoly(2-hydroxypropyl methacrylate) (PGMA47-PHPMAx)
diblock copolymers synthesized by RAFT aqueous dispersion polym-
erization at 70 °C. (b) DMF gel permeation chromatograms (vs poly-
(methyl methacrylate) standards) obtained for a series of six PGMA47-
PHPMAx diblock copolymers (and the corresponding PGMA47
macro-CTA) synthesized via RAFT aqueous dispersion polymerization
at 70 °C and 10 w/v % solids.
higher polydispersity was obtained (Mn = 145 500 g molꢀ1
,
Mw/Mn = 1.48; see Figure 2). This was attributed to small
amounts of a dimethacrylate impurity within the HPMA mono-
mer (0.11 mol % as judged by HPLC; see Supporting Informa-
tion Figure S3), which results in light branching of the PHPMA
chains. Silica column chromatography (CH2Cl2 eluent) was uti-
lized to reduce the dimethacrylate impurity content to below the
HPLC detection limit (<0.01 mol %). Using this purified HPMA
enabled the synthesis of a near-monodisperse G112-H1000 diblock
copolymer (Mn = 110 800 g molꢀ1, Mw/Mn = 1.18) (see Figure 2).
The G47-Hx copolymer particles were analyzed by TEM to
assess their morphology (see Figure 3). G47-H90 generated
exclusively spherical micelles (see Figure 3a). Increasing the
targeted PHPMA block length leads to a mixture of (mainly)
short, linear wormlike micelles and some remaining spherical
micelles for G47H115 (see Figure 3b). Targeting a block compo-
sition of G47-H130 leads to longer wormlike micelles, with some
worms now containing y-junctions or “branch points” (see
arrows in Figure 3c and Supporting Information Figure S5).
An increase of just 10 HPMA units results in the generation of
highly branched, wormlike micelle networks for G47-H140 (see
Figures 3d and Supporting Information Figure S5). G47-H150
forms an intriguing intermediate phase comprising highly
branched wormlike micelles that have partially coalesced to form
nascent bilayers (see Figures 3e and Supporting Information
Figure S5). Finally, a pure vesicle phase is observed when
targeting G47-H160 (see Figure 3f).
(Varian Module 230), a UV detector (Varian model 310), and a Waters
SymmetryShield RP18 3.5 μm, 4.6 ꢁ 150 mm HPLC column. The
batches of HPMA monomer were allowed to equilibrate at room
temperature for 30 min after removing from the freezer prior to opening
in order to suppress condensation. Samples were weighed in an
autosampler vial and dissolved in acetonitrile to give a concentration
of 6.0 mg/mL within 2 h of analysis. The eluent consisted of 0.10%
trifluoroacetic acid (TFA) in water and acetonitrile. The following
conditions were used: 0ꢀ5 min: 70% 0.10% TFA in water. 5ꢀ20 min:
70% 0.10% TFA in water to 100% acetonitrile. The UV detector
wavelength was 219 nm.
’ RESULTS AND DISCUSSION
The initial RAFT polymerization of GMA was conducted in
ethanol at 70 °C to generate a hydrophilic, near-monodisperse
PGMA47 macro-CTA (Mw/Mn = 1.13; Figure 1 and Supporting
Information Figure S1). After purification, this macro-CTA was
then utilized for the in situ RAFT aqueous dispersion polymer-
ization of HPMA to produce a series of well-defined G47-Hx
diblock copolymers at 10 w/v % (where G denotes GMA, H
1
denotes HPMA, and x = 90, 115, 130, 140, 150, and 160). H
NMR studies indicated relatively fast polymerization kinetics,
with >99% HPMA conversion being achieved within 2 h. Gel
permeation chromatography (GPC) studies indicated that near-
monodisperse diblock copolymers were obtained with very high
blocking efficiencies (Mw/Mn < 1.18) and minimal macro-CTA
contamination; see Figure 1b and Supporting Information
Figures S2ꢀ4). Furthermore, this formulation enables the
To assess the in situ structural evolution that ultimately leads
to vesicle formation by a G47-H200 copolymer at 10 w/v % and
70 °C, the polymerizing solution was periodically sampled for
1
TEM and H NMR analysis. The monomer conversion ver-
sus time plot provides a fascinating mechanistic insight (see
16583
dx.doi.org/10.1021/ja206301a |J. Am. Chem. Soc. 2011, 133, 16581–16587

Journal of the American Chemical Society
ARTICLE
Figure 3. Representative TEM images of the final particle morpholo-
gies (at >99% conversion and 10 w/v %) observed for a series of six
G47-Hx diblock copolymers, where x corresponds to (a) 90, (b) 115,
(c) 130, (d) 140, (e) 150, and (f) 160. Spherical, worm, and vesicular
nanostructures are observed simply by systematically increasing the
targeted length of the core-forming PHPMA block.
Figure 4. HPMA polymerization kinetics obtained for the targeted
G47-H200 diblock copolymer prepared via RAFT aqueous dispersion
polymerization at 70 °C and 10 w/v %. The five morphological regimes
as judged by TEM are as follows: molecularly dispersed copolymer
chains (M), spherical micelles (S), wormlike micelles (W), branched
wormlike micelles, jellyfish (BW & J), and vesicles (V). The inset shows
a semilogarithmic plot for a subset of these data, which confirms the
5-fold nucleation-induced rate enhancement observed on micellar
aggregation.
Figure 4). Three distinct regimes are observed. The first regime
(green line, see inset) suggests mild retardation, which is commonly
observed (but not fully understood) for RAFT polymerization of
methacrylates.34 The second regime is from 20 to 60 min (red
line); this corresponds to the formation of molecularly dispersed
copolymer chains and involves a modest increase in the slope of
the semilogarithmic plot. However, in the third regime, the rate
of polymerization increases by a factor of 5 after ∼1 h (blue line);
this corresponds to the onset of nucleation, since 20 nm micelles
are now observed by DLS. This reaction time corresponds to a
HPMA conversion of ∼45% (i.e., PHPMA90), which is consis-
tent with the critical block length previously reported for the
micellar aggregation of PHPMA-based diblock copolymers in
aqueous solution.35 We suggest that the unreacted HPMA
monomer enters these nascent micelles and solvates the growing
PHPMA chains, thus leading to a relatively high local concentra-
tion of monomer and hence the observed rate enhancement.
Each aliquot extracted for 1H NMR analysis was also analyzed
by TEM to assess the evolving predominant particle morphol-
ogies. As discussed above, no nanostructures were detected with-
in 1 h of the HPMA polymerization. At 65 min (46% conversion,
corresponding to G47-H92), spherical micelles are observed by
both DLS and TEM (see Figure 5a). A mixture of spherical
micelles and short wormlike micelles is formed after 70 min
(55%; G47-H110). On close inspection of Figure 5b, these short
worms are formed by the fusion of spherical micelles into dimers
and trimers. Previous studies of the aqueous self-assembly of
highly hydrophobic polybutadiene-based amphiphilic block co-
polymers indicate micellar fusion does not occur on experimental
time scales, resulting in a nonergodic system with kinetically
frozen structures.17,18 In contrast, micellar fusion has been re-
ported for several systems,36ꢀ38 but only when aided by either
cosolvent or added salt. In the present work, the PHPMA chains
are relatively solvated by both the HPMA monomer (at inter-
mediate conversions) and water (particularly at shorter PHPMA
chain lengths), facilitating the sphere-to-worm transition. This is
similar to the recent work of He et al.,38 who observed micellar
fusion at intermediate conversions during the RAFT alcoholic
dispersion polymerization of styrene using a poly(acrylic acid)
macro-CTA. Based on our own studies, we suggest that the
unreacted styrene monomer solvates these growing polystyrene
chains, thus facilitating micellar fusion. In order to confirm the
swelling of the PHPMA chains by water, a near-monodisperse
PHPMA200 homopolymer (Mn = 30 000 g molꢀ1, Mw/Mn =
1.13) was synthesized by RAFT in ethanol using the CPDB chain
transfer agent. This homopolymer was analyzed by differential
scanning calorimetry (DSC) to determine its glass transition
temperature (Tg). PHPMA200 rigorously dried under vacuum at
50 °C for 3 days exhibited a Tg of 95 °C. However, a thin film of
PHPMA200 that had been soaked in excess water for 10 days prior
to DSC studies exhibited a significantly reduced “hydrated Tg” of
47 °C due to water uptake. As the reaction temperature (70 °C)
is always higher than the “hydrated Tg’ of the core-forming block
(since the PHPMA chains within the cores will have some bound
water molecules), this suggests that these growing chains are
highly mobile, leading to ergodic aggregates being formed during
the polymerization. When the G47-H200 polymerization is sam-
pled at intermediate conversions, dilution of each extracted
aliquot causes any unreacted water-miscible HPMA monomer
to immediately diffuse into the aqueous continuous phase, thus
quenching the polymerization. Since rapid cooling to 20 °C also
occurs during sampling, the intermediate particle morphology is
rendered nonergodic (i.e., frozen) and is hence preserved for
TEM studies.
Both spherical micelles and linear wormlike micelles are still
present after 75 min (62%; G47-H123; see Figure 5c), but after
78 min (68%; G47-H131) wormlike micelles constitute the
primary morphology. However, at this point, the worms begin
to form branch points (primarily y-junctions; see arrows in
Figure 5d), due to a reduction in molecular curvature and
increased copolymer molecular weight.17 After 80 min (72%;
G47-H144), the number of worm branch points now increases
significantly, with worm clustering observed (see Figures 5e
and 6 and Supporting Information Figure S6). Figure 5f and
16584
dx.doi.org/10.1021/ja206301a |J. Am. Chem. Soc. 2011, 133, 16581–16587

Journal of the American Chemical Society
ARTICLE
Figure 5. Intermediate nanostructures observed during the sphere-to-worm and worm-to-vesicle transitions. Transmission electron micrographs
obtained for (a) spheres, (b) short worms, (c) long worms, (d) branched worms, (e,f) partially coalesced worms, (g) jellyfish, and (hꢀj) vesicles
generated in situ after various reaction times for a target G47-H200 diblock copolymer prepared by RAFT aqueous dispersion polymerization at 70 °C and
10 w/v % solids. Scale bars = 200 nm.
Figure 6. Suggested mechanism for the polymerization-induced worm-to-vesicle transformation during the synthesis of G47-H200 by RAFT aqueous
dispersion polymerization.
16585
dx.doi.org/10.1021/ja206301a |J. Am. Chem. Soc. 2011, 133, 16581–16587

Journal of the American Chemical Society
ARTICLE
Supporting Information Figure S6 show how these highly
branched structures undergo partial coalescence and develop
nascent bilayers with protruding “tentacles” (84 min; 75%;
G47-H150). This octopus-like morphology has been previously
reported for a binary mixture of two poly(ethylene oxide)ꢀ
polybutadiene diblock copolymers.17 The present work suggests
that this is an intrinsic intermediate morphology between worms
and bilayers formed by polymerization-induced self-assembly of
a single copolymer. After 87 min (78%; G47-H156), these octopi
structures then undergo partial wrap-up and form “jellyfish” (see
Figures 5g and 6 and Supporting Information Figure S6) with
surface pores and defects due to incomplete worm/bilayer fusion
(see arrows in Figure 5g and h). In essence, the hemispherical
jellyfish structures appear to be the final stage prior to vesicle
formation. The jellyfish “tentacles” undergo fusion after 90 min
(82%; G47-H164). This produces a predominantly vesicular
phase, but some vesicles do remain interconnected via residual
“tentacles” (see Figure 5h and i). However, as the surface pores
coalesce at higher HPMA conversions, defect-free vesicles are first
observed after 100 min (91%; G47-H182) (see Supporting Infor-
mation Figure S6f) and finally a pure vesicle phase is formed after
225 min (>99%; G47-H200) (see Figures 5j and 6).
The above observations are consistent with previous studies of
the worm-to-vesicle transition that occurs on rapidly diluting
polystyreneꢀpoly(acrylic acid) diblock copolymers in dioxane/
water mixtures with further water.39 A two-step transition was
reported, in which the rods/worms first flatten to form lamellae,
followed by wrap-up of the lamellae to produce vesicles. How-
ever, the additional intermediate morphologies observed in the
present work, which are only observed over a narrow range of
diblock compositions, significantly enhance our understanding
of the detailed mechanism of vesicle formation.
relatively low styrene conversions (typically < 70%) are obtained.
The gradual evolution of spheres to worms to vesicles was
observed in this earlier study, but none of the complex intermediate
morphologies observed herein (partially coalesced nascent bilayers,
jellyfish, etc.) that are essential for a more complete mechanistic
understanding have been previously reported.38,45,46 We believe that
these new insights have been achieved by extensive sampling of the
in situ HPMA polymerization, although it is also possible that
there are intrinsic differences between RAFT aqueous dispersion
polymerization and RAFT syntheses conducted under either aqu-
eous emulsion or alcoholic dispersion polymerization conditions.
’ CONCLUSIONS
In summary, polymerization-induced self-assembly via RAFT
aqueous dispersion polymerization of HPMA using a PGMA
macro-CTA leads to near-monodisperse diblock copolymers,
which form spheres, worms, and vesicles, depending on the
targeted block composition. The onset of micellar nucleation cor-
responds to an enhancement in the rate of polymerization (>99%
HPMA monomer conversion within 2 h), which suggests solva-
tion of the growing PHPMA chains by the unreacted HPMA
monomer. Moreover, close monitoring of the in situ HPMA
polymerization by TEM reveals a range of intermediate morphol-
ogies, which provide important mechanistic insights regarding
the sphere-to-worm and worm-to-vesicle transitions.
’ ASSOCIATED CONTENT
S
Supporting Information. Details of the HPMA mono-
b
mer purification, further details of the synthesis of high molecular
weight G112-H1000 copolymers, and additional TEM images.
This material is available free of charge via the Internet at http://
pubs.acs.org
For the structural evolution during the synthesis of G47-H200
,
the in situ morphologies obtained at a given time (and hence
mean DP for the growing PHPMA block) correspond closely to
the final post mortem morphologies observed at full conversion
when targeting the equivalent DP. For example, spherical mi-
celles are obtained when targeting G47-H90 (see Figure 3a),
which is the same morphology as that observed after 65 min (or
46% conversion, corresponding to an in situ block composition
of G47-H92; see Figure 5a) during the polymerization of G47-
H200. Similarly, close correspondence between the in situ block
compositions and the final post mortem compositions is also
observed for the wormlike micelles and vesicles (compare
Figures 3 and 5). This is an important insight, because it suggests
that the unreacted HPMA monomer does not affect the inter-
mediate morphologies that are observed during the kinetic
studies when targeting a final block composition/morphology
of G47-H200 vesicles.
’ AUTHOR INFORMATION
Corresponding Author
s.p.armes@sheffield.ac.uk
’ ACKNOWLEDGMENT
A.B. thanks the EPSRC (EP/E012949/1) for postdoctoral
funding. Dr. Svetomir Tzokov is thanked for preparing the
carbon-coated TEM grids.
’ REFERENCES
(1) Alberts, B.; Johnson, A.; Lewis, J.; Raff, M.; Roberts, K.; Walter,
P. Molecular Biology of the Cell, 4th ed.; Garland Science: New York,
2002.
Our aqueous dispersion polymerization formulation offers a
number of advantages for polymerization-induced self-assembly com-
pared to aqueous emulsion40ꢀ43 or alcoholic dispersion38,44ꢀ46
polymerization formulations reported in the literature. In the for-
mer case, the worm or vesicle phases are typically contaminated
with spheres and, although monomer conversions are usually
high, they do not exceed 95%. Moreover, relatively high copol-
ymer polydispersities are often observed,40ꢀ42 although a new
report describes rather better results for such emulsion polymer-
ization syntheses.43 In the latter case, Pan and co-workers38,44ꢀ46
have reported low polydispersities (Mw/Mn < 1.20) for the
polymerization of styrene in methanol and relatively pure phases
can be obtained (i.e., solely spheres, worms or vesicles), but
(2) Balmbra, R. R.; Clunie, J. S.; Goodman, J. F. Nature 1969,
222, 1159–1160.
(3) Israelachvili, J. N. Intermolecular & Surface Forces, 2nd ed.;
Academic Press: London, 1991; Vol. 1.
(4) Israelachvili, J. N.; Mitchell, D. J.; Ninham, B. W. J. Chem. Soc.,
Faraday Trans. 1976, 72, 1525–1568.
(5) Zhang, L. F.; Eisenberg, A. Science 1995, 268, 1728–1731.
(6) Won, Y. Y.; Davis, H. T.; Bates, F. S. Science 1999, 283, 960–963.
(7) Li, Z. B.; Kesselman, E.; Talmon, Y.; Hillmyer, M. A.; Lodge,
T. P. Science 2004, 306, 98–101.
(8) Cui, H. G.; Chen, Z. Y.; Zhong, S.; Wooley, K. L.; Pochan, D. J.
Science 2007, 317, 647–650.
(9) Wang, X. S.; Guerin, G.; Wang, H.; Wang, Y. S.; Manners, I.;
Winnik, M. A. Science 2007, 317, 644–647.
16586
dx.doi.org/10.1021/ja206301a |J. Am. Chem. Soc. 2011, 133, 16581–16587

Journal of the American Chemical Society
ARTICLE
(10) Pochan, D. J.; Chen, Z. Y.; Cui, H. G.; Hales, K.; Qi, K.; Wooley,
K. L. Science 2004, 306, 94–97.
(11) Jain, S.; Bates, F. S. Science 2003, 300, 460–464.
(12) Discher, D. E.; Eisenberg, A. Science 2002, 297, 967–973.
(13) Christian, D. A.; Tian, A. W.; Ellenbroek, W. G.; Levental, I.;
Rajagopal, K.; Janmey, P. A.; Liu, A. J.; Baumgart, T.; Discher, D. E. Nat.
Mater. 2009, 8, 843–849.
(14) Massignani, M.; LoPresti, C.; Blanazs, A.; Madsen, J.; Armes,
S. P.; Lewis, A. L.; Battaglia, G. Small 2009, 5, 2424–2432.
(15) Howse, J. R.; Jones, R. A. L.; Battaglia, G.; Ducker, R. E.;
Leggett, G. J.; Ryan, A. J. Nat. Mater. 2009, 8, 507–511.
(16) Blanazs, A.; Armes, S. P.; Ryan, A. J. Macromol. Rapid Commun.
2009, 30, 267–277.
(17) Jain, S.; Bates, F. S. Macromolecules 2004, 37, 1511–1523.
(18) Fernyhough, C.; Ryan, A. J.; Battaglia, G. Soft Matter 2009,
5, 1674–1682.
(19) Discher, B. M.; Won, Y. Y.; Ege, D. S.; Lee, J. C. M.; Bates, F. S.;
Discher, D. E.; Hammer, D. A. Science 1999, 284, 1143–1146.
(20) Bermudez, H.; Brannan, A. K.; Hammer, D. A.; Bates, F. S.;
Discher, D. E. Macromolecules 2002, 35, 8203–8208.
(21) Battaglia, G.; Ryan, A. J. J. Am. Chem. Soc. 2005, 127,
8757–8764.
(22) Du, J. Z.; Tang, Y. P.; Lewis, A. L.; Armes, S. P. J. Am. Chem. Soc.
2005, 127, 17982–17983.
(23) Blanazs, A.; Massignani, M.; Battaglia, G.; Armes, S. P.; Ryan,
A. J. Adv. Funct. Mater. 2009, 19, 2906–2914.
(24) Napoli, A.; Valentini, M.; Tirelli, N.; Muller, M.; Hubbell, J. A.
Nat. Mater. 2004, 3, 183–189.
(25) Qin, S. H.; Geng, Y.; Discher, D. E.; Yang, S. Adv. Mater. 2006,
18, 2905.
(26) LoPresti, C.; Lomas, H.; Massignani, M.; Smart, T.; Battaglia,
G. J. Mater. Chem. 2009, 19, 3576–3590.
(27) Hayward, R. C.; Pochan, D. J. Macromolecules 2010, 43, 3577–
3584.
(28) Lipic, P. M.; Bates, F. S.; Hillmyer, M. A. J. Am. Chem. Soc. 1998,
120, 8963–8970.
(29) Hamley, I. W.; Stanford, J. L.; Wilkinson, A. N.; Elwell, M. J.;
Ryan, A. J. Polymer 2000, 41, 2569–2576.
(30) An, Z. S.; Shi, Q. H.; Tang, W.; Tsung, C. K.; Hawker, C. J.;
Stucky, G. D. J. Am. Chem. Soc. 2007, 129, 14493–14499.
(31) Delaittre, G.; Save, M.; Charleux, B. Macromol. Rapid Commun.
2007, 28, 1528–1533.
(32) Rieger, J.; Grazon, C.; Charleux, B.; Alaimo, D.; Jꢀer^ome, C.
J. Polym. Sci., Part A: Polym. Chem. 2009, 47, 2373–2390.
(33) Li, Y.; Armes, S. P. Angew. Chem., Int. Ed. 2010, 49, 4042–4046.
(34) Moad, G.; Rizzardo, E.; Thang, S. H. Acc. Chem. Res. 2008, 41,
1133–1142.
(35) Madsen, J.; Armes, S. P.; Bertal, K.; MacNeil, S.; Lewis, A. L.
Biomacromolecules 2009, 10, 1875–1887.
(36) Denkova, A. G.; Mendes, E.; Coppens, M. O. J. Phys. Chem. B
2009, 113, 989–996.
(37) Burke, S. E.; Eisenberg, A. Langmuir 2001, 17, 6705–6714.
(38) He, W.-D.; Sun, X.-L.; Wan, W.-M.; Pan, C.-Y. Macromolecules
2011, 44, 3358–3365.
(39) Chen, L.; Shen, H. W.; Eisenberg, A. J. Phys. Chem. B 1999,
103, 9488–9497.
(40) Delaittre, G.; Dire, C.; Rieger, J.; Putaux, J.-L.; Charleux, B.
Chem. Commun. 2009, 2887–2889.
(41) Rieger, J.; Zhang, W. J.; Stoffelbach, F.; Charleux, B. Macro-
molecules 2010, 43, 6302–6310.
(42) Boisse, S.; Rieger, J.; Belal, K.; Di-Cicco, A.; Beaunier, P.; Li,
M. H.; Charleux, B. Chem. Commun. 2010, 46, 1950–1952.
(43) Zhang, X.; Boissꢀe, S.; Zhang, W.; Beaunier, P.; D’Agosto, F.;
Rieger, J.; Charleux, B. Macromolecules 2011, 44, 4149–4158.
(44) Huang, C.-Q.; Pan, C.-Y. Polymer 2010, 51, 5115–5121.
(45) Cai, W.-M.; Wan, W.-M.; Hong, C.-Y.; Huang, C.-Q.; Pan, C.-Y.
Soft Matter 2010, 6, 5554–5561.
(46) Wan, W.-M.; Pan, C.-Y. Polym. Chem. 2010, 1, 1475–1484.
16587
dx.doi.org/10.1021/ja206301a |J. Am. Chem. Soc. 2011, 133, 16581–16587