Seeded emulsion polymerization of block copolymer core-shell nanoparticles with controlled particle size and molecular weight distribution using xanthate-based RAFT polymerization

Article abstract of DOI:10.1021/ma049496l

The work described here has shown for the first time that well-defined core-shell nanostructures with controlled particle size and molecular weight distributions (in which the molecular weight distributions of both blocks are independently controlled using the same RAFT controlling agent) can be prepared in an aqueous environment. These types of nanoparticles can lead the way to more novel and interesting structures in applications for drug delivery, biosensors, and other devices. A poly(methyl methacrylate) (PMMA) seed was used to control the particle size distribution, in which styrene (STY) in the presence of the RAFT agent was polymerized under batch conditions to give polymer particles with an number-average molecular weight (M_n) of 7000 and polydispersity (PDI) of 2. Butyl acrylate (BA) was then polymerized under semibatch conditions into these particles to give an overall M_n of 20 000 and PDI close to 1.3 for the block copolymer, in which greater than 90% block purity was observed. The calculated PDI of the second block (PBA) was close to 1.4, which is lower than the theoretically determined value under batch conditions (PDI = 1.6). The results show that by slow monomer addition feed into the reactor the PDI can be reduced and that by using the advantage of "random coupling" between the two blocks the PDI of the final block copolymer is lower than either of the individual blocks.

Full text of DOI:10.1021/ma049496l

4
474
Macromolecules 2004, 37, 4474-4483
Seeded Emulsion Polymerization of Block Copolymer Core-Shell
Nanoparticles with Controlled Particle Size and Molecular Weight
Distribution Using Xanthate-Based RAFT Polymerization
†
,‡
Wilfr ed Sm u ld er s a n d Mich a el J . Mon teir o*
Laboratory of Polymer Chemistry (SPC), Eindhoven University of Technology, P.O. Box 513, 5600 MB
Eindhoven, The Netherlands, and School of Molecular and Microbial Sciences, Australian Institute of
Bioengineering and Nanotechnology, University of Queensland, Brisbane QLD 4072, Australia
Received March 14, 2004; Revised Manuscript Received April 19, 2004
ABSTRACT: The work described here has shown for the first time that well-defined core-shell
nanostructures with controlled particle size and molecular weight distributions (in which the molecular
weight distributions of both blocks are independently controlled using the same RAFT controlling agent)
can be prepared in an aqueous environment. These types of nanoparticles can lead the way to more
novel and interesting structures in applications for drug delivery, biosensors, and other devices. A poly-
(
(
methyl methacrylate) (PMMA) seed was used to control the particle size distribution, in which styrene
STY) in the presence of the RAFT agent was polymerized under batch conditions to give polymer particles
with an number-average molecular weight (M
n
) of 7000 and polydispersity (PDI) of 2. Butyl acrylate
of 20 000
(BA) was then polymerized under semibatch conditions into these particles to give an overall M
n
and PDI close to 1.3 for the block copolymer, in which greater than 90% block purity was observed. The
calculated PDI of the second block (PBA) was close to 1.4, which is lower than the theoretically determined
value under batch conditions (PDI ) 1.6). The results show that by slow monomer addition feed into the
reactor the PDI can be reduced and that by using the advantage of “random coupling” between the two
blocks the PDI of the final block copolymer is lower than either of the individual blocks.
In tr od u ction
Sch em e 1. O-Eth ylxa n th yl Eth yl P r op ion a te (I)
Controlled synthesis of molecular structures of nano-
scale dimensions has opened an exciting realm of
1
-3
applications ranging from molecular computers
to
miniature medical devices capable of entering individual
4
cells to carry out a variety of functions. Among various
approaches to “smart” molecular architectures capable
of self-assembly the use of block copolymers has proven
quite useful.5 Incorporating blocks of different nature
into one polymer chain provides a driving force for
-7
subsequent phase separation and supramolecular struc-
a particle, the second monomer is added and further
polymerized to form a block copolymer with a core-shell
structure. The major advantages of aqueous media over
bulk or solution are that the polymerization is carried
out in an environmentally friendly medium and in
principle can be used directly in biomedical applications,
it is a cheap process, it can be used for a broad range of
monomers and a wide range of experimental conditions,
the heat transfer is highly efficient, high conversions
with low monomer residuals can be reached, there are
no organic volatile compounds that are detrimental if
these nanoparticles are used for human applications,
and one can obtain high polymer solids (up to 60 wt %)
in a low-viscosity environment (which means the poly-
mer is easy to process). Out of all of the living tech-
niques, RAFT is ideally suited for emulsion systems
since the activating species (-S-C(Z)dS, Scheme 1) is
attached to the polymer chain end and will not partition
into the water phase. Thus, the number of radical per
particle is not affected, resulting in a controlled “living”
process.
ture formation.8
-10
Functionality can be incorporated
into one of the blocks such that the shell can be cross-
6
,10
linked
for use as possible drug delivery or the surface
can be modified with an oligomeric peptides sequence
for specific cell binding.4 The applications of such
nanostructures even include nanoreactors for the prepa-
ration of gold particles.11 Block copolymers of this type
and even more complex architectures (e.g., hyper-
branched, star) can be synthesized in bulk monomer and
solution with the recently developed “living” free-radical
polymerization techniques. These techniques include
1
2
nitroxide-mediated polymerization (NMP), atom trans-
fer radical polymerization (ATRP),1
3,14
and reversible
addition-fragmentation chain transfer (RAFT) polym-
1
5,16
erization.
Although the methodology of forming nanoparticles
through self-assembly of block copolymers in water has
proven useful, we used another approach by synthesiz-
ing the blocks directly in water using one of the living
radical techniques. Once the first block is made to form
Adopting the RAFT technique to emulsion polymer-
ization has been a long standing focus of our group.
Besides providing an ability to run “living” radical
polymerizations in water, a nonhazardous and nontoxic
reaction media, emulsion systems provide radical com-
†
Eindhoven University of Technology.
University of Queensland.
To whom correspondence should be sent. E-mail: m.monteiro@
‡
*
uq.edu.au.
1
0.1021/ma049496l CCC: $27.50 © 2004 American Chemical Society
Published on Web 05/21/2004

Macromolecules, Vol. 37, No. 12, 2004
Block Copolymer Core-Shell Nanoparticles 4475
partmentalization within growing particles. This re-
duces bimolecular termination, which is important for
maintaining high rates of polymerization and high block
purity. Although RAFT polymerization is relatively
straightforward in solution and bulk for the “highly”
active RAFT agents such as dithioesters, RAFT-derived
polymers have proven difficult to control via classical
ab initio (i.e., where surfactant micelles are used as
mircoreactors for polymerization) and seeded (where
prepolymer particles are used as the microreactors)
emulsion polymerizations.17 In most cases, there was
little control of the molecular weight distribution (MWD),
and concomitantly severe colloid destabilization was
found. These problems may stem from the high reactiv-
ity of the RAFT agents used, producing oligomeric
material in the water phase that could not transport to
the growing particles.1
7-19
These problems were over-
2
0,21
come by miniemulsions to make homopolymers
and
block copolymers20 with narrow MWDs and high block
purity. Miniemulsions avoid the need for transport of
monomer or RAFT agent to the particles since the locus
of polymerizations is within the stable monomer drop-
lets. However, miniemulsions by their very nature
produce a broad particle size distributions (PSDs)
ranging from 30 nm to over 1 µm. The PSD must be
narrow in order for particles to be useful in most
applications involving nanotechnology.
F igu r e 1. Cryo-TEM of the core-shell morphology stained
4
with RuO and PTA (core size: 50 nm) of a block copolymer of
PSTY-block-PBA/P(acetoacetoxyethyl methacrylate) prepared
using xanthates. The dark core consists of PSTY, and the light
2
5
shell consists of PBA/P(acetoacetoxyethyl methacrylate).
ethylxanthyl ethyl propionate, Scheme 1, where Ctr,RAFT
2
2
30
to styrene is 0.69 and to BA is 1.7 ) to prepare high
block purity poly(styrene-block-n-butyl acrylate) core-
shell nanoparticles with a controlled MWD (in which
the MWDs of both blocks can be independently con-
trolled with the use of the same controlling agent) and
PSD. Previous work with xanthates to synthesize block
copolymers relied on the ab initio emulsion polymeri-
zation of the first monomer to determine the resulting
PSD. In this work, polymer particles (seed) of prede-
termined number and size distribution will be used to
control the number of particles and breath of the particle
size distribution for polymerizations carried out under
monomer feed conditions. The number of seed particles
was chosen as to avoid any secondary particle formation.
The use of slow monomer addition feed rates reduces
the monomer concentration in the particles, and for low
Ctr,RAFT agents should theoretically result in narrow
MWDs of homopolymers and block copolymers com-
pared to batch polymerizations as will be shown by our
computer simulations. This methodology to prepare
block copolymer in a dispersed (aqueous) media of well-
defined core-shell nanostructures would obviate the
need for self-aggregation and allow by simply changing
the size of the seed and controlling the amounts of two
monomers, block copolymer nanostructures with a
desired particle size to be prepared. The amount of
initial seed material in the final block copolymer nano-
structure should be small, on the order of a few percent,
to minimize the effect of the seed material on the final
properties.
We then shifted our attention to xanthates as RAFT
agents as they were less studied due to their low Ctr,RAFT
value, resulting in a broad MWD.22-24 The main advan-
tage of using these RAFT agents is that, to our knowl-
edge, they are currently the only “living” agents that
can be successfully used in a classical ab initio emulsion
(
waterborne) polymerization for the synthesis of novel
2
5
nanostructures. Homopolymers made by this method
gave a controlled and predictable (albeit broad, PDI is
2
for styrene and 1.6 for n-butyl acrylate polymeriza-
tions) MWDs and broad particle size distributions
PSDssin the range from 20 to 120 nm, which is still
(
much narrower than the miniemulsion experiments),
and were subsequently used in a second stage polym-
erization to prepare block copolymers with block purity
as high as 90%.26 The high block purity was achieved
by feeding the monomer into the reaction vessel slowly
over a 5 h period. Cryo-TEM demonstrated that these
nanostructures have a core-shell morphology with the
core consisting of polystyrene (PSTY) and the shell
consisting of poly(butyl acrylate) (PBA).25 To explain the
fact that a core-shell morphology together with high
block purity can be prepared, we postulated that the
xanthate RAFT agents used in these studies were
surface active, such that the -S-C(OEt)dS moiety on
the dormant polymer chain end resides at the surface
of the particles.2
5,27
A close look at the structure of the
xanthate shows that one of its resonance structures
2
8
leads to an ionized (water-soluble) structure. Using
this phenomenon, we were also able to prepare reactive
Exp er im en ta l Section
Ma ter ia ls. Styrene (STY, 99%), n-butyl acrylate (BA, 99%),
and methyl methacrylate (MMA, 99%) were purchased from
Aldrich and purified of inhibitor by passing through an
inhibitor-removal column (Aldrich). 2,2-Azobis(isobutyro-
nitrile) (AIBN, 99%) was purchased from Aldrich and recrys-
tallized twice from ethanol. Toluene (99%), sodium dodecyl
sulfate (SDS, 98%), sodium persulfate (SPS, >98%), and
2
5
core-shell nanostructures consisting of copolymers
(
Figure 1), where a reactive functional group (aldehyde)
is incorporated into the shell for subsequent cross-
linking with diamines. The resulting cross-linked films
gave novel mechanical tensile properties which could
be tuned depending upon the location of the functional
groups in the shell.29 Other factors such as surface
NaHCO
received.
3
(>99.5%) were purchased from Aldrich and used as
tension of the polymers also aid in core-shell formation.
The goal of this work is to use the surface-active
properties of the low reactive xanthate RAFT agents (O-
Syn th esis of O-Eth ylxa n th yl Eth yl P r op ion a te (I).
O-Ethylxanthyl ethyl propionate (Scheme 1, I) was synthesized
according to the literature procedure23 by adding potassium

4
476 Smulders and Monteiro
Macromolecules, Vol. 37, No. 12, 2004
O-ethyldithiocarbonate (101.4 g, 0.633 mol, Merck, 99%) to a
mixture of ethyl 2-bromopropionate (102 g, 0.563 mol, Merck,
Ta ble 1. Rea ction Con d ition s a n d Resu lts of a Block
Cop olym er iza tion of Styr en e a n d n -Bu tyl Acr yla te w ith
RAF T in Solu tion , in Wh ich Styr en e Is P olym er ized F ir st
in th e P r esen ce of I
9
9%) dissolved in ethanol (1 L) at 0 °C under a nitrogen
atmosphere. The mixture was stirred for 4 h at 0 °C in the
absence of light. Water (1 L) was added, and the product was
extracted by a 1:2 mixture of diethyl ether and pentane. The
solvent was removed, and the remaining ethyl 2-bromopro-
pionate distilled off under vacuum. I was obtained at >99%
recipe
conditions
result
First Step: Styrene Homopolymerization
0
.016 g of AIBN
70 °C
21 h
V
46% conversion
Mn ) 2010
1
10.020 g of styrene
2
1
purity according to H NMR.
.116 g of I (Scheme 4.2)
0.002 g of toluene
PDI ) 1.44
Solu tion P olym er iza tion s. Styrene Homopolymerization
precipitation in polystyrene (PS1)
methanol
-
3
(
PSTY-Xanthate). AIBN (0.016 g, 4.31 × 10 M) was added
to a degassed reaction vessel containing STY (10.02 g, 4.26
M), I (2.116 g, 0.46 M), and 11.56 mL of toluene. Oxygen was
removed from the mixture by bubbling through with nitrogen.
The reaction mixture was placed in an oil bath at 70 °C and
left to polymerize for 21 h. During this time samples were
taken, and conversion (by gravimetry) and the MWD (by SEC)
were determined. The conversion after 21 h reached 46%, and
the polymer consisted of a number-average molecular weight
Second Step: n-Butyl Acrylate Block Copolymerization
0
6
0
1
.043 g of AIBN
.412 g of n-butyl acrylate 23.5 h
.508 g of PS1
60 °C
88% conversion
Mn ) 14640
PDI ) 2.15
5.080 g of toluene
a mass ratio of 4 to 1 of BA to PSTY was reached. In three
separate experiments the feed rates were 0.02, 0.04, and 0.08
(M
n
) of 2000 and a polydispersity (PDI) of 1.44 according to
GPC.
Poly(styrene-butyl acrylate) Block Copolymerization. AIBN
-
1
mL min . The resulting polymer particles had an average
particle diameter of 284 nm, which is close to the calculated
value of 265 nm.
-
2
(
0.043 g, 1.06 × 10 M) was added to a degassed reaction
GP C. The dried polymer was dissolved in tetrahydrofuran
vessel containing BA (6.412 g, 2.50 M), PSTY-xanthate (0.508
(THF, Biosolve) to a concentration of 1 mg/mL. The solution
-
2
g, 1.03 × 10 M), and 17.43 mL of toluene. Oxygen was
removed from the mixture by bubbling through with nitrogen.
The reaction mixture was placed in an oil bath at 60 °C and
left to polymerize for 23.5 h. The conversion after 23.5 h
reached 88%, and the polymer consisted of a number-average
was filtered over a 0.2 mm PTFE syringe filter. Analysis was
carried out using a Waters model 510 pump, a Waters model
WISP 712 autoinjector, a model 410 refractive index detector,
and a model 486 UV detector (at 254 nm). The columns used
were a PLgel guard (5 µm particles) 50 × 7.5 mm precolumn,
followed by two PLgel mixed-C (5 µm particles) 300 × 7.5 mm
columns in series (which were maintained at 40 °C for
analysis). The columns used separate polymers in the molec-
ular weight range between 500 and 2 million with high
resolution. THF was used as an eluent (flow rate 1.0 mL/min),
and calibration was done using polystyrene standards (Poly-
molecular weight (M
.15 according to GPC.
Syn th esis of Block Cop olym er La tex P a r ticles w ith
Con tr olled P SD. Poly(methyl methcarylate) (PMMA) Seed.
n
) of 14 600 and a polydispersity (PDI) of
2
-
2
MMA (90.8 g, 0.91 mol), SDS (3.0 g, 1.04 × 10 mol), NaHCO
3
(
0.3 g), and water were mixed into a 250 mL jacketed glass
reactor equipped with a reflux condenser. The mixture was
stirred overnight with a magnetic stirrer under an argon
atmosphere at room temperature. The reactor was then heated
6
mer Laboratories; M ) 580-7.1 × 10 ). Data acquisition was
performed using waters Millenium 32 (v3.05) software. The
molecular weights of the block copolymers were in some cases
corrected with Mark-Houwink parameters given in the
-3
to 80 °C and the initiator, SPS (0.3 g, 1.26 × 10 mol), added.
The polymerization was stopped after 4 h and extensively
dialyzed to remove residual surfactant, initiator, monomer,
and other low molecular weight compounds. Dialysis was
carried out by pouring the latex into a semipermeable cellulose
membrane tubing (Spectra/Por 4) and dialyzed against deion-
ized water. The dialyzate was changed regularly. The resulting
polymer particles had an average particle diameter of 57 nm,
and with a solids content of 7.31 wt %. This seed is used to
control the PSD.
literature:31 a ) 0.716 and K ) 11.4 × 10
-5
dL g for
-1
-
5
-1
polystyrene and a ) 0.700 and K ) 12.2 × 10 dL g for
poly(n-butyl acrylate).
HP LC. HPLC analyses were performed using an Alliance
Waters 2690 separation module, equipped with a PL-EMD 960
ELD detector and a 2487 Waters dual UV detector, operating
at 254 and 280 nm. The block copolymers were analyzed using
a Zorbax Si column, thermostated at 35 °C using a gradient
from pure n-heptane (Biosolve) to THF (Biosolve) in 40 min.
Typically 10 µL of dried sample dissolved in THF at 5 mg/g
THF was injected.
Polystyrene-Xanthate (PSTY-Xanthate) Latex Particles.
PMMA seed (27.0 g, 7.31 w% solids), STY (31.52 g, 0.302 mol),
-
3
SDS (0.33 g, 1.26 × 10 mol), NaHCO
3
(0.058 g), and water
(
143.53 g) were mixed into a 250 mL jacketed glass reactor
Resu lts a n d Discu ssion
equipped with a reflux condenser. The mixture was stirred
overnight with a magnetic stirrer under an argon atmosphere
at room temperature. The reactor was then heated to 60 °C
Deter m in a tion of Ctr ,RAF T for BA to P STY-Xa n -
th a te fr om Solu tion P olym er iza tion s. Low molec-
ular weight polystyrene was synthesized in the presence
of a low transfer constant RAFT agent, i.e., O-ethyl-
xanthyl ethyl propionate (I, Scheme 1), at 70 °C (see
Table 1 for experimental details). The amount of dead
chains was minimized by keeping the initiator concen-
tration very low (1 mol % of the amount of RAFT agent).
The polymer formed was precipitated in methanol,
filtered, and further washed with methanol. The latter
was done in order to remove residual monomer and
RAFT agent, thereby excluding their influence on the
second block polymerization. Since the polymerization
did not proceed to high conversion, a considerable
amount of unreacted RAFT agent was still present
because of the low chain transfer constant of the RAFT
-
4
and the initiator, SPS (0.063 g, 2.6 × 10 mol), added. The
polymerization was stopped after 22.5 h and dialyzed to
remove residual surfactant, initiator, monomer, and other low
molecular weight compounds. Conversion reached 96% after
2
n
2.5 h, and the polymer had an M of 7000 and PDI of 2.0
according to GPC. The resulting polymer particles had an
average particle diameter of 157 nm, which is close to the
calculated value of 149 nm. The solids content was 13.1 wt %.
These particles were then used to make latex particles
consisting of block copolymers.
Poly(styrene-butyl acrylate) Block Copolymer Latex Par-
ticles. PSTY-xanthate latex (33.2 g, 13.1 wt % solids), BA (2
-
2
-4
g, 1.9 × 10 mol), SDS (0.162 g, 5.6 × 10 mol), NaHCO
3
(
0.024 g), and water (68.4 g) were mixed into a 250 mL jacketed
glass reactor equipped with a reflux condenser. The mixture
was stirred for 18 h with a magnetic stirrer under an argon
atmosphere at room temperature. The reactor was then heated
2
2
agent with styrene. The number-average molecular
-
5
weight (M ) of the polymer formed was 2011, and the
to 60 °C and the initiator, SPS (0.0236 g, 9.9 × 10 mol),
added. After 20 min, BA was then slowly fed into the reactor
using a Metrohm Dosimat syringe pump and was stopped until
n
polydispersity (PDI) was 1.44. This low PDI (compared
to a PDI of about 2 expected for transfer-dominated STY

Macromolecules, Vol. 37, No. 12, 2004
Block Copolymer Core-Shell Nanoparticles 4477
F igu r e 2. Number-average molecular weight (closed squares), M
n
, and polydispersity (open circles) of the solution block
-
2
copolymerization of n-butyl acrylate (2.5 M), polystyrene-xanthate (1.03 × 10 M; M
n
) 2010, PDI ) 1.44), initiated with AIBN
-
2
(
1.06 × 10 M) in 17.43 mL of toluene at 60 °C. The solid line (s) represents the theoretical evolution predicted by eq 3. PREDICI
simulations (Appendix 1) of the polydispersity profiles for different transfer constants are shown. Ctr,RAFT ) 1.0 (- - -); Ctr,RAFT
)
4
3
-1 -1 31
1
1
.31 (- ‚ -); Ctr,RAFT ) 1.73 (‚ ‚ ‚). Parameters used in the simulations are as follows: k
p
) 2.29 × 10 dm mol
s
,
k
d
) 9.8 ×
-
6
-1 33
34
8
3
-1 -1
0
s
,
initiator efficiency f is estimated to be 0.6, and k
t
) 4 × 10 dm mol
s , since this value fitted the conversion time
data.
homopolymerizations) is probably the result of the
precipitation, in which the lowest molecular weight
polymer is selectively removed. The PSTY obtained was
then used in a solution polymerization with BA. The
molecular weight and PDI vs conversion of the second
step as obtained by GPC analysis using a refractive
index (RI) detector are shown in Figure 2.
In the ideal case, where no homopolymer of BA is
formed by assuming negligible termination, the theo-
retical Mn is given by
simulation package, Scheme 2), the Mn vs conversion
profile is identical to eq 1, suggesting that these side
reactions are not significant under our reaction condi-
tions. In other words, simulations validate eq 1 for our
system, and therefore eq 1 will be used to describe
changes in Mn with conversion from subsequent emul-
sion polymerizations. The difference between the ex-
perimental and theoretical lines remains unclear, but
it is speculated that this difference could be due to the
error in determining low molecular weights by GPC that
are close to the exclusion limit.
mB
A
A
The PDI initially increased to a value of about 4.5 and
then decreased toward 2.15 (at 88% conversion). This
is not surprising, since both the newly formed block
copolymer and the starting block combine to give the
PDI. Figure 2 shows the computer simulations (see
Scheme 2 for the kinetic scheme) of the PDI at different
transfer constants for BA to polystyrene-xanthate, i.e.,
1.0, 1.3, and 1.73 (using the experimental concentrations
M ) Mn + x
Mn
(1)
n
mA
in which MnA is the number-average molecular weight
of the first block (A), x is the fractional conversion of
the second monomer (B), mA is the total mass of the
polymer A, and mB is the total initial mass of monomer
B. The validity of this equation for low reactive RAFT
agents is because polymeric RAFT agents are used, in
which the number of polymer chains is already fixed at
the start of the reaction, and monomer units are simply
added to each chain with time. This equation becomes
useful in an emulsion system where consumption of
monomer and other substances are described in terms
of mass instead of concentration. The data in Figure 2
are based on PSTY standards and, therefore, quantita-
tively might contain an error for block copolymers.
Figure 2 shows that Mn increased linearly from 2011
to approximately 14 000 with conversion. The molecular
weights are much lower than theoretically expected
from eq 1. Correction of the molecular weights with the
Mark-Houwink parameters of PBA only gives slightly
higher molecular weights. The differences in molecular
weight between theory and experiment might arise from
side reactions from termination of radicals and chain
transfer reactions to monomer and solvent, which are
not taken into account in eq 1. However, even when
termination and chain transfer reactions are taken into
4
3
-1 -1 31
-6
and kp ) 2.29 × 10 dm mol
s
,
kd ) 9.8 × 10
-
1 33
34
s
, the initiator efficiency f is estimated to be 0.6
and kt ) 4 × 10 dm mol
8
3
-1 -1
s
, since this value fitted
the conversion time data). The transfer constant for this
RAFT agent lies between 1 and 1.73 based on simula-
tions. The linear increase in Mn with conversion and
the similar PDI profile to the simulated data supports
the formation of block copolymers. Additional evidence
for the formation of block copolymer is obtained from
the GPC traces. Using a UV detector on the GPC, the
shift of the PSTY can be monitored. If the molecular
weight of the starting PSTY dormant species grows by
the addition of BA, then the molecular weight distribu-
tion as observed from UV detection should also increase.
This is due to the fact that PBA does not have an
extinction coefficient at 254 nm, and thus is considered
“invisible”. Figure 3a indeed shows a shift of the UV
GPC trace to higher molecular, indicating that the
starting PSTY-xanthate has grown in molecular weight.
This is strong evidence for the formation of a block
copolymer.
3
2
account using a computer simulations (PREDICI

4
478 Smulders and Monteiro
Macromolecules, Vol. 37, No. 12, 2004
Sch em e 2. P REDICI Sim u la tion of a Block
Cop olym er iza tion
F igu r e 3. (a) Normalized GPC UV (254 nm) traces of the
solution block copolymer of n-butyl acrylate (2.5 M) polymer-
-
2
ized, polystyrene-xanthate (1.03 × 10 M), initiated with
-
2
AIBN (1.06 × 10 M) in 17.43 mL of toluene at 60 °C as a
function of conversion, and (b) the starting and final MWDs
(
at 88% conversion) are compared by the RI detector and the
UV detector at 254 nm.
has previously been described by Goto et al.35,36 It was
derived that
Since these traces have been normalized, it also shows
d
dt
S0
that the amount of initially present dormant PSTY
decreases. The fact that at 88% conversion most of the
GPC signal shifted toward the high molecular weight
side indicates that most of the PSTY starting material
has been converted to block copolymer, and this type of
block copolymer synthesis is quite effective. In Figure
ln
S
M0
Ctr,RAFT
)
(2)
d
dt
ln
M
in which S0 is the amount of transfer agents at t ) 0, S
is the amount of transfer agent, M0 the amount of
monomer at t ) 0, and M the amount of monomer. If
the natural logarithm of S0/S and the natural logarithm
of M0/M are plotted vs time, then the transfer constant
is given by the ratio of the slopes of these plots. Here
the amount of transfer agent, in this case PSTY-
xanthate, is estimated from the height of the low
molecular weight side of the UV GPC trace. The amount
of monomer can be obtained from the conversion mea-
surements (see Figure 4).
The transfer constant derived from the data in Figure
4 for BA to PSTY-xanthate is 1.04. This value is an
underestimation and in fact will be higher. The amount
of PSTY-xanthate is estimated from the low molecular
weight side of the GPC UV trace, which will also contain
some block copolymer as a result of the broadness of
the high block copolymer peak. Therefore, the amount
3
b the starting and final MWDs are shown. Besides the
clear shift toward higher molecular weight of the MWD
obtained from the UV detector, this figure also shows
that the final MWDs from UV and RI do not completely
overlap as a result of unreacted PSTY-xanthate and
“
dead” PSTY. Figure 3b also indicates that the amount
of high molecular weight BA homopolymer is small,
suggesting that the resulting polymer contains a high
percentage of block copolymer.
The UV traces above can be used to estimate the
Ctr,RAFT for BA radicals to dormant PSTY chains (i.e.,
with a xanthate as the end group, denoted here as
“
PSTY-xanthate”). The UV traces can be normalized
since the amount of PSTY in the reactor is constant,
and thus the area below the UV signal over a designated
elution time span indicates the amount of material in
the corresponding molecular weight span. The method

Macromolecules, Vol. 37, No. 12, 2004
Block Copolymer Core-Shell Nanoparticles 4479
F igu r e 4. Determination of the chain transfer constant of n-butyl acrylate toward polystyrene-xanthate using the method of
Goto. The transfer constant, Ctr,RAFT, is obtained from conversion measurements and the GPC traces in Figure 2 using eq 2. The
-
2
-2
reaction mixture consisted of n-butyl acrylate (2.5 M), polystyrene-xanthate (1.03 × 10 M), initiated with AIBN (1.06 × 10 M)
in 17.43 mL of toluene at 60 °C. This results in a Ctr,RAFT of 1.04.
F igu r e 5. (a) Simulations of the chain length distribution P(M), (b) the weight distribution W(M), (c) the GPC distribution, and
(
d) the polydispersity vs chain extension for an ideal block copolymerization (i.e., all initially present polymer chains add the
same number of monomer units in the second stage) in which the starting block has an M of 7000 and a polydispersity of 2
Flory-Schultz distribution, eq 3). In (d), the polydispersity vs chain extension of an ideal block copolymerization is compared to
n
(
4
a PREDICI simulation of a batch block copolymerization (Scheme 2) using Ctr,RAFT ) 10 and a very low initiator concentration.
of PSTY-xanthate will be overestimated, resulting in a
underestimated transfer constant. We30 found that the
transfer constant of BA with I was close to1.7. The true
transfer constant for BA to PSTY-xanthate therefore lies
between 1 and 1.73 (see simulation in Figure 2). The
synthesis of polystyrene-block-poly(n-butyl acrylate)

4
480 Smulders and Monteiro
Macromolecules, Vol. 37, No. 12, 2004
Ta ble 2. Recip es a n d P a r ticle Sizes of th e Sem iba tch Em u lsion Block Cop olym er iza tion s of Styr en e a n d n -Bu tyl
Acr yla tea
recipe
conditions
result
First Step: PMMA Seed Latex Preparation
water 942 g
SDS 3.0 g
NaHCO3 0.3 g
SPS 0.3 g
MMA 90.8
80 °C
particle size dv ) 57 nm
solid contents after dialysis: 7.31 wt %
PMMA seed
4 h
extensive dialysis afterward
Second Step: Polystyrene-Xanthate Latex Preparation
water 143.53 g
SDS 0.33 g
60 °C
96% conversion
22.5 h
Mn ) 7000, PDI ) 2.0
NaHCO3 0.058 g
PMMA seed 27.0 g
SPS 0.063 g
dv ) 157 nm (calculated 149 nm)
solid contents after dialysis: 13.1 wt %
polystyrene-xanthate seed
STY 31.52 g
Third Step: Block Copolymer Latex Preparation
stir for 18 h at room
water 68.4 g
0.02 mL/min, dv ) 284 nm
0.04 mL/min, dv ) 284 nm
0.08 mL/min, dv ) 281 nm
(dv(calculated) ) 265 nm)
polystyrene-block-poly(n-butyl acrylate)
latex
SDS 0.162 g
(without initiator), raise
NaHCO3 0.024 g
polystyrene-xanthate seed 33.2 g
SPS 0.0236 g
temperature to 60 °C, add
initiator, wait 20 min and
feed residual monomer (15.7 mL
BA 2.0 g
) 14.1 g) at 0.02, 0.04 and 0.08 mL/min,
respectively.
a
The following polymer densities were used:33 1.05 g/mL (PSTY), 1.15 g/mL (PMMA), and 1.03 g/mL (PBA).
using a xanthate is therefore effective, and the transfer
constant of BA to PSTY-xanthate is of the same order
of magnitude as to I.
Figure 5 shows the chain length distributions P(M),
weight distributions W(M), log weight distribution w(log
M), and polydispersity vs Mn of an ideal block copolym-
erization obtained through computer simulations for
varying degrees of chain extension. The initially present
block has an Mn of 7000 and a polydispersity of 2 and
is obtained from eq 3.
Syn th esis of P olystyr en e-block-poly(n -bu tyl acr y-
la te) in Em u lsion u n d er Sem iba tch Con d ition s
w ith a Con tr olled MWD a n d P SD. Low-polydisper-
sity block copolymers using RAFT agent I cannot be
obtained under batch conditions because of its low
Ctr,RAFT to both BA30 and styrene, being approximately
Figure 5 shows that an ideal block copolymerization
can lead to narrow molecular weight material, even
though the starting material has a PDI of 2. For
instance, a chain extension from Mn ) 7000 to Mn )
22
1
.7 and 0.7, respectively. It is generally found that the
PDI is broad and close to 1.4 for BA and 2 for STY.
However, M u¨ ller et al. have shown that it is theoreti-
cally possible to produce low-polydispersity material
with low transfer constant species if the monomer
concentration is kept low.37 Krstina et al. have experi-
mentally confirmed this for a radical system in solution
by the production of low-polydispersity blocks using
1
4 000 leads to a decrease in the PDI from 2 to 1.25
(Figure 5d), whereas a chain extension to Mn of 25 000
leads to a PDI of 1.08.
Ideal block copolymerization conditions can also be
approached if the transfer constant in the RAFT process
is very high. In Figure 5d, a PREDICI simulation of the
3
8
macromonomers under semibatch conditions. If the
monomer concentration is low enough, then ideally each
growing radical will add no more than one monomer
unit before the radical activity is transferred to another
chain. This means that all chains grow simultaneously,
resulting in a low PDI. In the case of an ideal block
copolymerization of BA on PSTY-xanthate, all initially
present PSTY-xanthate chains will add an equal num-
ber of monomer units to form the new block copolymer.
In terms of a number distribution (concentration vs
chain length), the starting MWD will be translated over
the chain length axis toward a higher molar mass.
Importantly, this means that an ideal living polymer-
ization can theoretically be imitated by performing the
block copolymerization under semibatch conditions.
Id ea l Block Cop olym er iza tion . An ideal block
copolymerization (all initially present chains of the first
block will add the same number of monomer units) will
result in a translation of the chain length distribution
of the initial block over the chain length axis. The chain
length distribution of such a polymer can mathemati-
cally be approximated by the Flory-Schultz distribu-
polydispersity vs chain extension is shown for a batch
block copolymerization with a transfer constant of 104
and a low initiator concentration. It appears that a
system having a transfer constant of this order of
magnitude, which is not unrealistic for high reactivity
36
RAFT agents, behaves like an ideal block copolymer-
ization, but these highly reactive RAFT agents have
been shown to lead to colloid destabilization in emulsion
polymerizations. Interestingly, the use of low reactive
xanthate systems also gave the same PDI profile (Figure
5
d, solid line) as that for the high reactive agents,
suggesting that in principle the use of low active RAFT
agents will produce the same MWD as the highly active
ones. The theory above suggests that xanthates, which
can be successfully used in classical ab initio emulsion
systems, can also be used to produce polymer of a
narrow molecular weight distribution.
There are a number of other complicating factors that
may affect the PDI. These include (a) side reactions, e.g.
bimolecular termination, transfer to monomer, or poly-
mer; (b) under feed conditions where the feed rate is
slow enough to produce low-PDI polymer, the polymer-
ization rate will be infinitely slow such that the reaction
cannot be performed in a practical way; and (c) not all
chains have an equal probability to grow at the same
rate if a broad PSD seed is used. The latter consider-
3
9
tion (eq 3).
Mi
P(M ) ) exp -
(3)
i
(
Mn)

Macromolecules, Vol. 37, No. 12, 2004
Block Copolymer Core-Shell Nanoparticles 4481
ation is due to the fact that per unit of volume all
particle sizes contain the same number of polymer
chains. The smaller particles have a larger surface area
per volume unit and, therefore, have a higher prob-
ability to be entered by an aqueous phase radical.
Exp er im en ta l vs Th eor y. As we have shown in
earlier work,25 the easiest approach to synthesize block
copolymers in emulsion is first to polymerize styrene
via an ab initio emulsion conditions in the presence of
the xanthate and, once this step is complete, to further
polymerize BA to make the block copolymer. However,
control of the PSD was very much dependent upon the
amount of xanthate used in the ab initio step; the
greater the amount of xanthate, the lower the average
particle size. The key to control the PSD in this work is
to use a seed with the desired particle size, number, and
distribution. The seed will be used as the sites for
polymerization, and therefore as polymerization pro-
ceeds, the average partyicle size will increase but the
breath and number of particles will remain constant.
The next step is the seeded polymerization of styrene
in the presence of xanthate, I, followed by polymeriza-
tion of BA to make block copolymer latex particles of
the same particle number as the initial seed.
The formation of block copolymers, when polystyrene
is the starting block, can be observed due to the shift of
the UV GPC trace toward higher molecular weights.
Therefore, in the following emulsion experiments sty-
rene was polymerized first in the presence of a xanthate
and a poly(methyl methacrylate) (PMMA) seed (57 nm
average volume diameter with a low polydispersity of
1
.09). The seed was used so that the resulting PSTY
latex formed would have a controlled particle number
and particle distribution. PMMA was chosen as the seed
since PMMA does not absorb at 254 nm and is thus
considered “UV-invisible”. Hence, by using a UV detec-
tor, the MWD of the newly formed PSTY can be readily
obtained. The content of PMMA by the end of the
polymerization is only a few percent (4-6%) of the PSTY
and upon subsequent block formation will decrease to
less than 1% and is for our purpose considered negli-
gible. The polydispersity of PSTY was close to 2 with
an Mn of 7000 as predicted (Table 2).
1
7
As our computer simulations have shown, it is theo-
retically possible to produce low-PDI block copolymer
using RAFT agents with a low transfer constant, even
though the first block has a relatively high PDI of about
n
F igu r e 6. (a) M vs overall conversion and (b) polydispersity
vs overall conversion for semibatch emulsion block copolymer-
izations of PSTY-xanthate latex (33.2 g, 13.1 w% solids), SDS
-
4
(
0.162 g, 5.6 × 10 mol), NaHCO
3
(0.024 g), water (68.4 g),
-
5
and SPS (0.0236 g, 9.9 × 10 mol) at 60 °C at BA feed rates
2
. To test whether low PDI block copolymer can be
-1
of 0.02, 0.04, and 0.08 mL min . The solid line represents the
obtained using xanthates, a series of experiments were
performed in which BA was fed to a PSTY-xanthate
latex with a predetermined PSD (recipes are given in
Table 2). The final BA to polystyrene mass ratio was 4
to 1 in all experiments, and the final solid contents of
the latex was 20%. The polymerization was performed
at 60 °C, the surfactant concentration was kept below
the critical micelle concentration, and the initiator
theoretically expected M
particles had an average particle diameter of 284 nm, which
is close to the calculated value of 265 nm.
n
(eq 1). The resulting polymer
expected molecular weights, it is obvious that the
experimental values are much lower. Figure 6b shows
that the PDI decreases with increasing conversion from
2 to between 1.3 and 1.62, again depending upon the
feed rate.
(sodium persulfate) concentration was 1 mM, such that
the amount of dead chains as a result of bimolecular
termination was kept low. The consumption of monomer
in the semibatch polymerizations was found not to be
strictly controlled by the feed; i.e., the measured overall
conversion is not equal to the monomer feed profiles.
When the monomer feed rate was the highest, Mn
decreased more rapidly away from theory than the other
slower feed rates. The cause may be due to transfer to
monomer. Figure 6b shows that the PDI decreased with
conversion for all the feed rates used. At the end of the
polymerization relatively low PDI material was pro-
duced, in which the PDI was 1.3 at the lowest feed rate
and 1.6 at the highest feed rate. At lower monomer feed
rates (where PDI is 1.3) the system is closer to the ideal
block copolymerization conditions with a block purity
Figure 6 shows the Mn and PDI vs conversion. Figure
a shows that the experimental Mn increases with
6
conversion from 7000 to between 15 000 and 20 000
depending upon the feed rate, but if the experimental
molecular weights are compared with the theoretically

4
482 Smulders and Monteiro
Macromolecules, Vol. 37, No. 12, 2004
F igu r e 7. Log weight molecular weight distribution of the starting polystyrene-xanthate (M ) 2010, PDI ) 1.44, solid line) and
n
the final block copolymer of a semibatch emulsion polymerization of PSTY-xanthate latex (33.2 g, 13.1 w% solids), SDS (0.162 g,
-
4
-5
5
.6 × 10 mol), NaHCO
3
-
(0.024 g), water (68.4 g), SPS (0.0236 g, 9.9 × 10 mol), and BA (which was fed in over 790 min at a
1
feed rate of 0.02 mL min ) at 60 °C, obtained by both RI detection (dashed line) and UV detection (dotted line).
Sch em e 3. Cor e-Sh ell (Block Cop olym er ) Na n ostr u ctu r es Syn th esized w ith Na r r ow P a r ticle Size Distr ibu tion
a n d Molecu la r Weigh t Distr ibu tion Usin g Low Active Xa n th a te (RAF T) Agen ts
as high as 90% (Figure 7), and it was calculated from
the “random coupling” methodology40 that the PDI of
the PBA block was close to 1.4. The block purity was
confirmed by HPLC by monitoring the loss of STY-
xanthate with monomer conversion.
Con clu sion
The highly desired preparation of core-shell mor-
phology with controlled molecular weight (in which the
MWDs of both blocks were independently controlled
using the same RAFT agent) and with a controlled
particle number and size distribution has been prepared
using surface-active RAFT agents (xanthates) of low
reactivity under semibatch conditions. A PMMA seed
was used to control the number and distribution of the
particles, in which STY in the presence of the RAFT
agent was polymerized under batch conditions to give
particles containing PSTY with an Mn of 7000 and PDI
of 2. Butyl acrylate was then polymerized under semi-
batch conditions into these particles to give an overall
Mn of 20 000 and PDI close to 1.3. This results in a PDI
of the second block to be calculated to be close to 1.4,
which is lower than the theoretical value of 1.6 for a
batch polymerization. The results show that a slow
monomer feed during the polymerization decreased the
PBA polydispersity and that through the advantage of
“random coupling” of the blocks the overall polydisper-
sity of the final block copolymer is lower than either
block. The other advantage of this methodology is that
high-purity blocks are formed (>90%). Although the
starting PSTY block had a PDI of 2 and the second PBA
Figure 7 shows that the UV MWD has shifted
completely toward a higher molecular weight, which
indicates that block copolymer has formed. From the
matching overlays of the UV and RI MWDs it can be
concluded that the block copolymer formed is of a high
purity. The average particle diameter dv (Table 2) for
all the various feed rates are close to 284 nm, which is
close to the theoretically calculated dv of 265 nm. The
difference is probably due to the error in the size
analysis. We have shown that core-shell nanoparticles
with a controlled PSD and where the second block has
a much narrower MWD can be made using low reactive
RAFT agents (see Scheme 3). The formation of a core-
shell morphology has been previously described with
evidence from transmission electron microscopy (see
Figure 1). Even without the use of a surface active
RAFT agent, a core-shell morphology will result when
BA is polymerized onto a PSTY particle. Therefore, it
is believed that since this system is very similar to those
previously described the same core-shell morphologies
will result.

Macromolecules, Vol. 37, No. 12, 2004
Block Copolymer Core-Shell Nanoparticles 4483
block a PDI less than 1.4, it is envisaged that polymer-
izing the first block (i.e., styrene and RAFT agent in a
PMMA seed) under feed conditions can form even
narrower blocks.
Rizzardo, E.; Thang, S. H. Macromolecules 1998, 31, 5559.
16) Matyjaszewski, K., Ed. Living/ Controlled Radical Polymer-
ization; ACS Symposium Series 768; American Chemical
Society: Washington, DC, 2000.
(
(
17) Monteiro, M. J .; Hodgson, M.; de Brouwer, H. J . Polym. Sci.,
A block copolymer latex with a desired particle size
can be prepared by simply changing the size and
number of the seed latex particles (e.g., PMMA) and
controlling the amounts of two monomers. However, a
few considerations should be taken into account. First,
conditions should be chosen such that no secondary
particle formation takes place, which in practice means
that the particle number of the PMMA seed has to be
high enough. Second, the amount of initial PMMA seed
material in the final block copolymer latex should be
small, on the order of a few percent, so as to have no
effect on the final latex properties.
The work described here has shown for the first time
that well-defined nanostructures of controlled particle
size and molecular weight distributions can be prepared
in an aqueous environment, leading the way to more
novel and interesting structures in applications for drug
delivery, biosensors, and other devices.
Part A: Polym. Chem. 2000, 38, 3864.
(18) Tsavalas, J . G.; Schork, F. J .; de Brouwer, H.; Monteiro, M.
J . Macromolecules 2001, 34, 3938.
19) Prescott, W. W.; Ballard, M. J .; Rizzardo, E.; Gilbert, R. G.
G. Aust. J . Chem. 2002, 55, 415.
20) de Brouwer, H.; Monteiro, M. J .; Tsavalas, J . G.; Schork, F.
J . Macromolecules 2000, 33, 9239.
(
(
(21) Lansalot, M.; Davis, T. P.; Heuts, J . P. A. Macromolecules
2
002, 35, 7582.
(
22) Adamy, M.; van Herk, A. M.; Destarac, M.; Monteiro, M. J .
Macromolecules 2003, 36, 2293.
(
23) Charmot, D.; Corpart, P.; Michelet, D.; Zard, S.; Biadatti, T.;
Rhodia Chemie, 1998; Vol. WO 9858974 (Chem. Abstr. 1999,
130, 82018).
24) Charmot, D.; Corpart, P.; Adam, H.; Zard, S. Z.; Biadatti, T.;
Bouhadir, G. Macromol. Symp. 2000, 150, 23.
25) Monteiro, M. J .; de Barbeyrac, J . Macromolecules 2001, 34,
(
(
4
416.
(26) Monteiro, M. J .; Sjoberg, M.; Gottgens, C. M.; van der Vlist,
J . J . Polym. Sci., Part A: Polym. Chem. 2000, 38, 4206.
27) Smulders, W.; Gilbert, R. G.; Monteiro, M. J . Macromolecules
(
2
003, 36, 4309.
(
28) Moad, G.; Chiefari, J .; Chong, Y. K.; Krstina, J .; Mayadunne,
Refer en ces a n d Notes
T. A.; Postma, A.; Rizzardo, E.; Thang, S. H. Polym. Int. 2000,
4
9, 993.
29) Monteiro, M. J .; de Barbeyrac, J . Macromol. Rapid Commun.
002, 23, 370.
30) Smulders, W. Technische Universiteit Eindhoven, Eindhoven,
002.
(
1) Xia, Y.; Rogers, J . A.; Paul, K. E.; Whitesides, G. M. Chem.
(
(
(
Rev. 1999, 99, 1823.
2
(
(
(
2) Barrett, C. R. Mater. Res. Soc. Bull. 1993, 3.
3) Service, R. F. Science 1996, 273, 1834.
4) Liu, J .; Zhang, Q.; Remsen, E. E.; Wooley, K. L. Biomacro-
molecules 2001, 2, 362.
2
31) Beuermann, S.; Paquet, D. A.; McMinn, J . H.; Hutchinson,
R. A. Macromolecules 1996, 29, 4206.
(
(
(
5) Stupp, S. I.; LeBonheur, V.; Walker, K.; Li, L. S.; Huggins,
K. E.; Keser, M.; Amstutz, A. Science 1997, 276, 384.
6) Thurmond, K. B.; Kowalewski, T.; Wooley, K. L. J . Am. Chem.
Soc. 1996, 118, 7239.
7) Remsen, E. E.; Thurmond, K. B.; Wooley, K. L. Macromol-
ecules 1999, 32, 3685.
(
(
32) Wulkow, M. Macromol. Theory Simul. 1996, 5, 393.
33) Brandrup, J ., Immergut, E. H., Grulke, E. A., Eds. Polymer
Handbook, 4th ed.; J ohn Wiley & Sons: New York, 1999.
(34) Buback, M.; Huckestein, B.; Kuchta, F.-D.; Russell, G. T.;
Schmid, E. Macromol. Chem. Phys. 1994, 195, 2117.
(35) Fukuda, T.; Goto, A. Macromol. Rapid Commun. 1997, 18,
683.
(
(
8) Harada, A.; Kataoka, K. Science 1999, 283, 65.
9) Butun, V.; Billingham, N. C.; Armes, S. P. J . Am. Chem. Soc.
1
998, 120, 11818.
10) Butun, V.; Billingham, N. C.; Armes, S. P. J . Am. Chem. Soc.
998, 120, 12135.
11) Liu, S.; Weaver, J . V. M.; Save, M.; Armes, S. P. Langmuir
002, 18, 8350.
(36) Goto, A.; Sato, K.; Tsujii, Y.; Kukuda, T.; Moad, G.; Rizzardo,
E.; Thang, S. H. Macromolecules 2001, 34, 402.
(37) M u¨ ller, A. H. E.; Zhuang, R.; Yan, D.; Litvenko, G. Macro-
molecules 1995, 28, 7335.
(
(
(
(
1
2
(38) Krstina, J .; Moad, G.; Rizzardo, E.; Winzor, C. L.; Berge, C.
12) Georges, M. K.; Veregin, R. P. N.; Kazmaier, P. M.; Hamer,
G. K. Trends Polym. Sci. 1994, 2, 66.
13) Kato, M.; Kamigaito, M.; Sawamoto, M.; Higashimura, T.
Macromolecules 1995, 28, 1721.
14) Wang, J .-S.; Matyjaszewki, K. Macromolecules 1995, 28, 7901.
15) Chiefari, J .; Chong, Y. K.; Ercole, F.; Krstina, J .; Le, T. P.
T.; Mayadunne, R. T. A.; Meijs, G. F.; Moad, G.; Moad, C. L.;
T.; Fryd, M. Macromolecules 1995, 28, 5381.
(
(
39) Gilbert, R. G. Emulsion Polymerization: A Mechanistic
Approach; Academic: London, 1995.
40) Tanaka, T.; Omoto, M.; Inagaki, H. J . Macromol. Sci., Phys.
(
(
1
980, B17, 211.
MA049496L