Morphology of Latex Particles Formed by

Poly(methy1Methacrylate)-Seeded Emulsion
Polymerization of Styrene

IWHAN CHO and KYUNG-WOO LEE, Department of Chemistry, Korea

Advanced Institute of Science and Technology, P.O. Box 150 Chongyangni,
Seoul 131, Korea.

Synopsis
Poly(methy1 methacrylatebpolystyrene composite particle latexes were prepared by
poly(methy1methacrylatehseeded emulsion polymerization of styrene employing batch, swell-
ing-batch, and semibatch methods. The changes in particle morphology taking place during
the polymerization reaction were followed by electron microscopy. Anchoring effect exerted
by ionic terminal groups introduced by ionic initiator was found to be the main factor in
controlling the particle morphology. The polymer particles obtained by oil-soluble hydrophobic
initiators such as azobisisobutyronitrile and 4,4’-azobis44-cyanovalericacid) gave the inverted
core-shell morphology. Water-soluble hydrophilic initiator, K2&08,also gave the inverted core-
shell morphology. However, in this case the occurrence of the halfmoonlike, the sandwichlike,
and the core-shell morphologies were also observed depending upon the polymerization con-
ditions. The distribution of terminal S O T groups on the surface area of polystyrene particles
could be controlled by initiator concentration and polymerization temperature. Viscosity of
polymerization loci dictated the movement of polymer molecules, thus causing the unevenness
of particle shape and phase separation at high viscosity state. Viscosity was controlled by the
styrene/poly(methyl methacrylate) ratio, the addition of a chain transfer agent or a solvent
which is common to polystyrene and poly(methy1 methacrylate).

INTRODUCTION
Control of latex particle morphology is important for many practical
applications.’ Latex particles of different morphological features have been
prepared from seeded emulsion polymerization techniques. This technique
has been employed for various kinetic2 and morphology studies involving
homo- and heterogeneous polymer pair systems.
Hughes and Brown3were the first to investigate the physical properties
of heterogeneous composite polymer particles, and Paxbn4 followed and
confirmed the results employing the polystyrene (PS)core/poly(methyl
methacrylate) (PMMAbshell system with soap titration method. Dickie and
his co-workers5 studied the particle morphology of heterogeneous acrylic
latex particles (a pair of glassy and rubbery polymers) by transmission
electron microscopy and minimum film-forming temperature method. In a
series of works Okubo and his co-workers61’ showed that heterogeneous
composite polymer particles could change its particle morphology from con-
fettilike to raspberrylike and/or void-containing, depending upon the mo-
lecular weights of polymers, viscosity of polymerization loci, polymerization
condition, and the difference in hydrophilicity of two polymers. Lee and his
co-~orker’”’~studied the effects of stage ratio, molecular weights of poly-

Journal of Applied Polymer Science, Vol. 30, 1903-1926 (1985)

@ 1985 John Wiley & Sons, Inc. CCC 0021-8995/85/0519034$04.00
1904 CHO AND LEE

mers, and the hydrophilicity of polymers on the particle morphology and

summarized well the various types in the particle morphology. Recently,
Min and his co-workers15have shown that the morphology of poly(buty1
acrylate) (PBAkPS composite latex particle was affected by the adhesion
effect exhibited by the byproduct graft copolymer distributed at the inter-
face of two different polymers in a particle.
Selective staining of double bonds in polymer backbones by Os0416makes
transmission electron microscopy a useful tool for the morphological study
of polymer blends. Recently, Trent and his co-workers17have found that
RuO, could stain PS region selectively in PELPMMA blends. Thsy also
reported that RuO, could stain certain other polymers which were not
stained by OsO,.
Most of the works reported hitherto are limited to the morphology and
the physical properties of glassy-rubbery polymer pair s y ~ t e m s , 3 . ~ J l - ~ ~ J ~
the reason being that the prepared composite polymers might have certain
improved properties. Lacking in the literatures are the studies on the
change of the particle morphology which are taking place particularly dur-
ing the polymerization reactions. In the present work we have taken a
glassy-glassy polymer pair system (PS-PMMA) and followed the change of
latex particle morphology. Various polymerization parameters were taken
into account, and the change of the particle morphology during the poly-
merization reaction was monitored by taking transmission electron micro-
graphs (TEMs) with Ru0,-staining method and scanning electron
micrographs (SEMs). For the clearer interpretation the monodispersed
PMMA seed latexes were used.

EXPERIMENTAL
Materials. Reagent grade styrene and methyl methacrylate (MMA)were
distilled under reduced pressure. The middle fraction was cut and stored
in a refrigerator. Sodium dodecylbenzene sulfonate (SDBS, Siponate DS-10
from Alcolac Chemical Co.), and Triton X-100 (polyoxyethylene isooctyl-
phenylether from Rohm and Haas Co.) were used without further purifi-
cation. Water was redistilled from basic KMnO, solution. Other chemicals
were analytical grade reagents and used without further purification.
Polymerization. All the polymerizations were performed in l-L flasks
under N2 purging. Each flask was equipped with a paddle-type stirrer, a
reflux condenser, a sampling tube, a dropping funnel, and thermocouples.
Polymerization temperature was controlled by electrically regulated heat-
ing mantle and recorded with Linear 141 recorder. Agitation speeds were
120 rpm for MMA polymerization and 200 rpm for PMMA-seeded styrene
polymerization. N2 purging was started with the charge of ingredients into
the flask and continued to the peak time. In a general procedure all the
ingredients were first charged and then the system was heated to the re-
quired temperature before the addition of initiator solution. Because of the
large heat of polymerization the temperatures of PMMA polymerization
could not be controlled rigorously. After the peak time the reaction systems
were heat-treated at 95°C for 2 h and then cooled with an ice-water bath
to ambient temperature.
 PMMA-PS COMPOSITE LATEX 1905
In semibatch polymerizations, the monomer was added dropwise to the
preheated reaction mixture, the addition rate being kept at 0.6 mL/min.
To follow the conversions and the morphological change of latex particles
taking place during the polymerization process, aliquots of the latex (20-
30 mL) were taken out at certain time intervals and hydroquinone solution
was added to stop the polymerization. Total solid contents (TSCs) were
determined gravimetrically.
Particle Morphology. The morphology of latex particles were studied
by three kinds of electron microscopic methods; latex TEMs, latex SEMs,
and Ru0,-stained thin section TEMs. General procedures for electron mi-
croscopic work are as follows:
(1)The apparent cross-sectional shapes of latex particles were observed
by latex TEMs. A latex sample diluted (ca. 50 ppm) with water was dropped
on the collodion-coated copper grid and then dried in open air. The residual
monomer and other volatile matters were removed under vacuum before
the micrographs were taken.
(2) For SEMs the dried latex grids prepared as above were placed in an
ion sputter (JFM-1000)and then coated with Au. The coating thickness was
about 50 A. The 3-dimensional shapes of particles were then observed from
those specimens with SEM mode, and the desired photographs were taken.
(3)The internal morphology of latex particles was determined as follows:
The latex (TSC, ca. 10%)which had been dialyzed usually for more than 7
days using a Visking cellophane tube was spread on a clean slide glass and
dried in air. Because of the high glass transition temperature (PS, 100°C;
PMMA, 105"C), the particles did not form film but small flat fragments.
The fragments were dried overnight at 60°C in a vacuum oven and then
transferred into BEEM capsules carefully, Epon 812 epoxy resin was used
for embedding and cured at 60°C for 18 h. The cured blocks were sectioned
with LKB Ultratome V ultramicrotome (thickness, ca. 900 A). These thin
sections were mounted on copper grids, dried in air, and stained over 0.1%
aqueous RuOl solution for 12 h. The stained sample grids were dried under
vacuum overnight before taking electron micrographs.
(4) Shadow casting: Being different from PS particles the PMMA particles
were shrunk by electron beam during the process, of taking TEMs. There-
fore, for the determination of precise particle size, the dried PMMA latex
grids, prepared as above [(l)], were shadowed with silver metal in Vacuum
Evaporater (JEM4X, manufactured by JEOL Co.). In a common procedure
the sizes of 100-300 particles were measured, and the values averaged.
The weight average D,, and the number average D, diameters and the
dispersity U of PMMA latex particles were calculated as follows:

D, = (Ln,Di/lni)

U = DJD,
where ni is the number of particles in the latex having diameter Di.
A
1906 CHO AND LEE

standard PS latex (0, = 2538 A, standard deviation = 31 A), supplied by

Dow Chemical Co. through the courtesy of Dr. D. I. Lee, was used as an
internal standard for size calibration.
TEMs and SEMs of these samples were obtained with a JEOL 100-CX
Electron Microscope. Accelerating voltages of electron beam were 60-100
kV for TEMs and 20 kV for SEMs.
Intrinsic Viscosity. Latexes were coagulated with hot acidic methanol
(1 vol % concn H2S04/CH30H).The precipitates were filtered and then
washed with a large volume of water. Thus obtained solid polymers were
reprecipitated with THF-methanol and dried in a vacuum oven at 60°C to
constant weight. Intrinsic viscosities of PMMAs were determined at 30
0.YC with a Cannon-Fenske No. 50 Viscometer (concn, 0.5 g/dL; solvent,
toluene) and calculated by the one-point method.
Swelling of Latex Particles. The swelling rates and the equilibrium
swelling ratio of PMMA latexes with styrene were measured under the
similar conditions employed for the polymerizations. Aliquots of collected
latex samples during the swelling experiments were centrifuged at 3000
rpm for 5 min. Then the styrene-swollen latexes were withdrawn by a
hypodermic syringe, and TSC was measured gravimetrically. Those TSC
values at 60°C and 80°C for the latex samples were quenched with an ice-
water bath and then the same procedures were applied. The swelling ratio
was calculated from TSCs before and after the swelling.
Other Measurements. Surface tension and interfacial tension were mea-
sured at room temperature by du Nouy ring methods. pH values of latexes
were determined by pH meter at room temperature after the heat treat-
ment.

RESULTS AND DISCUSSION

Preparation of Seed PMMA Latexes

Polymerization conditions and the properties of the resulting PMMA
latexes are summarized in Table I. Because of the difficulty in the control
of the polymerization temperature, the particle size of the resulting latexes
did not turn out as expected, even when the similar polymerization con-
ditions were employed. The easy radical desorption caused by chain transfer
agents decreased the particle size, exhibiting little effect on the molecular
weights. The high surface tension of PMMA latexes showed that the surface
of PMMA latex particles were not completely covered with emulsifiers. At
any event, as shown in Table I, the particle size distribution of each PMMA
latex obtained in the present study showed good monodispersity.

Seeded Emulsion Polymerization

Emulsion polymerization takes place in complicated mechanisms, and
the rate of polymerization and the morphology of latex particles can be
affected by many reaction parameters. In the present study the particle
nucleation stage of emulsion polymerization was bypassed by employing
seeding technique and new particle nucleation was minimized. In some cases
 TABLE I
Polymerization Conditions and Properties of Seed Poly(methy1 Methacrylate) Latexes
Sample name PMMA-1 PMMA-2 PMMA-3 PMMA-4 PMMA-5 PMMA-6 PMMA-7 g
Ingredients" ?
HzO 180 180 180 180 180 180 180 23
MMA 100 100 100 100 100 100 100
Triton X-100 3.01 3.0 3.0 3.0 3.0 3.0 3.0
SDBS 0.30 0.30 0.30 0.15 0.30 0.30 0.15
8g
KaSaOa 0.5 0.5 0.49 0.5 0.5 0.5 0.5 8
CHBr3 0 0 0 0 0.5 2.3 0
Properties
0. (A) 1560 1610 1700 3180 1330 1170 2570
U 1.003 1.003 1.002 1.001 1.002 1.004 1.001 g
PH 2.15 2.20 2.90 2.35 2.40 2.50 2.95
Surface tension (dyn/cm) 44.5 46.4 45.3 42.3 48.9 50.6 43.6 4x
Intrinsic viscosity 1.78 1.75 1.74 1.82 1.74 1.60 2.11
a Quantities are in weight parts.
1908 CHO AND LEE

the new particle nucleation was observed, but its amount was too small to
affect the system.

Swelling Time
Smith and EwartIg assumed that during the emulsion polymerization the
monomer concentration in polymerizing particles was independent of con-
version as long as monomer droplets were present in the system. Morton
and his co-workersmreported the fast diffusion rate of toluene into PS latex
particles. In those cases the latex particles were small (0, I loo0 A>.
However, Chung-Li and his co-workersZ1showed that when larger seed par-
ticles were used, the diffusion of monomer into the seed PS particles could
be a slow process.
Matsumoto and his co-workers6claimed that the particle morphology was
affected by the swelling state. Min and his ~o-workers'~ reported that the
swelling state affected the formation of grafted copolymers and the long-
term morphology change of PBA-PS composite latex particles.
In the present study we made an attempt to follow closely the effects of
swelling state on the particle morphology. As shown in Table 11, PMSd
and -6 showed the same rate of polymerization in spite of the difference in
the swelling time. And in Figure 1the two latex particles appeared as the

Fig. 1. Effect of swelling time on the particle morphology. (a) and (b) are latex TEMs and
(c) and (d) are TEMs of RuOI-stainedthin sections. PMS5 [(a) and (c); swelling time, 1 h] and
PMM [(b) and (d) swelling time, 4 days]. In (c) and (d) the dark part in a particle is the P S
rich region and the bright part is the PMMA-rich region.
 TABLE I1
Polymerization Conditions and Results of PMMA-Seeded Emulsion Polymerization of Styrene at 60°C~
Experimental no. PMS5 PMS-6 PMS-13 PMS16 PMSl2
iz
Ingredients
?
cd
Water 100 100 100 97.5 95
tn
PMMA (speciedb 5 (-1) 5 (-1) 5 (-2) 5 (-2) 5 (-2) 8
Styrene 15 15 15 15 15 iz
KzSzO, 0.1 0.1 0.1 0.1 0.1 cb
KOH (0.1M - - 0.5 2.5 5 %
Conditions 2
Swelling time (days) - 4 - - - M
PH 2.8 3.0 3.0 5.7 10.6
Rate of polymerization ( X 102 g/min) 3.75 3.87 3.63 3.47 3.50 r
a Quantities are in weight parts. 2x
Weight of seed latex is on dry base.
1910 CHO AND LEE

3- .C
A---
___---

same “halfmoonlike” morphology. It was obvious from these results that

the particle morphology and the rate of polymerization were not affected
by the swelling time at least in the system presently investigated.
The swelling rates of PMMA latexes (PMMA-1 and -4) with styrene at
room temperature are shown in Figure 2, and the equilibrium swelling
ratios (the weight of swollen styrenekhe weight of PMMA) of them are
given in Table 111. The equilibrium swelling ratio decreased with the in-
creasing polymerization temperature and increased with the particle size.
In our batch polymerization procedure the elapsing time starting from the
ingredients charge to the initiator addition was about 1h, and the swelling
ratio of PMMA-1 (in PMS5 polymerization) just before the initiator addition
(at 60°C) was 1.76 g/g. Although the equilibrium swelling ratio of PMMA-
1 at room temperature was 2.79 g/g, it decreased to 1.51 g/g at 60°C. Re-
gardless of the length of the swelling time at room temperature, the swelling
ratios at polymerization temperature were similar in all of our polymer-
ization systems. Therefore, it can be stated that the particle morphology
and the rate of polymerization were not affected by the swelling time.

TABLE I11
Equilibrium Swelling Ratios of PMMA-1 and -4 at Various Temperatures”
Room temp 60°C 80°C
PMMA-1 2.79 1.51 1.37
PMMA-4 3.20b 2.41 1.80
a Swelling Test Conditions: water, 95; PMMA (dry weight), 5; styrene, 16; hydroquinone, 0.1

(weight part)
b Maximum swelling ratio of this system, not equilibrium swelling ratio.
 PMMA-PS COMPOSITE LATEX 1911

pH Effects
In the PS-seeded emulsion polymerizations of styrenezzpH of the reaction
system did not affect the polymerization, when the value was greater than
4.0. It has been knownB that the rate of decomposition of KzSz08remains
almost unchanged through pH range between 3 and 12, but below 3 the
rate increases sharply. In PMMA-styrene systems the rate of polymeriza-
tion (Table 11; PMS-13, -16, and -12) and the particle morphology were not
affected by pH of the polymerization medium in this pH range, as expected.

Effects of Initiators
Three initiators, KaZO8,azobisisobutyronitrile (AIBN), and 4,4'-azobis-(4
cyanovaleric acid) (ABCVA), were used for the study. K2Sz08-initiatedpoly-
merizations under different conditions are summarized in Table IV. The
representative TEMs showing the change in apparent latex particle shape
and morphology taking place during the polymerization under different
reaction conditions are shown in Figures 3-5. The latex particles being in
the polymerization process were swollen with styrene, but the particles
became a monomer-free state when subjected to the procedures for TEM.
In a colloidal particle the external mixing of dispersing medium usually

Fig. 3. Change in apparent latex shape during the polymerization of PMS19 at 60°C. The
monomer conversionswere (a) 10.5%, (b)21.3%,(c) 52.1%, and (d) 93.5%. The volume growth
ratio (5.54/1) of particles between (a) and (d) is greater than the calculated value (3.43/1). It
was caused by the volume shrinkage of PMMA latexes under the electron beam in electron
microscope. Halfmoonlike morphology is clearly seen in (c).
 TABLE IV
Effects of Different Initiators under Different Conditions upon Polymerizations
PMS19 -24 -30 -31 -49 -52 -54 -55
~~~ ~ ~~ ~~ ~

Ingredients
Water 95 90 95 95 98 100 96 96
PMMA (species) 5 (-3) 5 (-3) 5 (-4) 5 (-4) 5 (-7) 5 (-7) 5 (-1) 5 (-4)
Styrene 15 15 15 15 15 15 15 15
KOH (0.1N) 5 10 5 5 2 4 4
Q*
AIBN 0.1
ABCVA 0.1
3F
K2S208 0.1 0.1 0.1 0.1 0.4 0.003 M
Conditions M
Polymerization
temp (“C) 60 80 80 60 80 60 80 80
a
Quantities are in weight parts.
When AIBN was used, the final latex had larger amount of coagulum (ca. 2-4 parts), but the coagulum could be minimized (ca. 0.1 part) by sufficient
premixing of AIBN with seed latex added with styrene solution (5 parts) at room temperature.
 PMMA-PS COMPOSITE LATEX 1913

Fig. 4. Change in apparent latex shape during the polymerization of PMS-24 at 80°C. The
monomer conversions were (a) 12.1%, (b) 30.0%, (c) 56.6%, and (d) 99.9%.The void-containing
particles are seen in 6).

does not perturb the internal phase of the particle. Therefore, the poly-
merizing particles were in a quasistatic state, and the particle morphology
was determined by the thermodynamic factors such as polymer incompat-
ibility and interfacial tension. The kinetic factors such as mobility of poly-
mer and viscosity of polymerization loci could also be operative.
Styrene was a common solvent to PS and to PMMA, but at high polymer
concentration (the maximum weight fraction of styrene in these systems
was 0.75) the PS- and PMMA-styrene phase started to ~ e p a r a t e ~as~ ,soon
~'
as styrene polymerization took place. In such a process the graft copolymer
was formed during the seeded polymerization3J5 and distributed at the in-
terfacial area of PS-PMMA, thus contributing to stabilize the separated
domains in each particle.
Lee and Ishikawa14 and Matsumoto et al.'j showed that in the hydrophilic-
hydrophobic polymer pair systems the morphology of particles depended
upon hydrophilicity, stage ratio, molecular weight, viscosity, and polymer-
ization methods. And when hydrophobic polymers were polymerized with
hydrophilic polymer seeds it was shown that the second polymers were
surrounded by seed polymer. This so-called "inverted core-shell" morphol-
ogy caused by the higher hydrophilicity of seed PMMA was well observed
also in our system, when oil-soluble (AIBN) or water-soluble but less hy-
1914 CHO AND LEE

a) b) C\
Fig. 5. Change in apparent latex shape during the polymerization of PMS31 at 60°C. The
monomer conversions were (a)3.7%, (b)31.4%, and (c) 92.5%.The sandwichlikeparticles are
seen in (b).

drophilic (ABCVA) initiator were used. As shown in Figure 6, the hydro-

phobic PSrich domain grew inside the relatively more hydrophilic PMMA-
rich matrix and formed “inverted core-shell” morphology. The same trend
was observed in the ABCVA-initiated polymerization.
In the contrast with what was described above we have obtained inter-
esting results with ionic K,S208-initiated polymerization systems. In this
case the particle morphology was little affected by the difference in the
hydrophilicity of PMMA and PS, but rather depended upon the initiator
concentration and polymerization temperature. When the initiator concen-
tration was low (PMS-54, concentration 0.003 g/dL), the PSrich domain
grew inside of the PMMA-rich matrix in a particle as illustrated in Figure
7. However, this formation of inverted core-shell morphology was not as
clear as in the case of PMS-49. In case of PMS3l (Fig. 8) the PS-rich domain
grew first at the tips of PMMA particles, and gradually the PS-rich domains
surrounded the PMMA-rich regions as the polymerization proceeded, thus
the PMMA-rich region becoming sandwiched by PS-rich domains. We here-

-0.3~ a b - 0.3~ -03p C

Fig. 6. TEMs of Ru04-stainedthin sections. Change in internal particle morphology of
PMS-49 during the polymerization at 80°C with AIBN. The monomer conversions were (a)
9.1%,(b) 28.0%, and (c) 86.0%.The inverted core-shell (PScore, PMMA-shell)morphology is
seen in (a) and (b).
 PMMA-PS COMPOSITE LATEX 1915

Fig. 7. Change of internal morphology of PMS54 during the polymerization (K&O, con-
centration, 0.003 g/dL). The monomer conversions were (a) 22.1% and @) 47.8%. The inverted
core-shell morphology can be deduced.

after call this morphology “sandwichlike.” In the case of PMS30 the P S

rich domain started in the similar way as PMS31, as shown in Figure 9.
However, in this case the PS-rich domains gradually covered the PMMA-
rich region. And finally the “core-shell” morphology emerged. This core-
shell morphology was also observed in PMS-24 (Fig. 10) and PMS-52 (TEMs
are not given) with a little difference. In the cases of PMS5, -6, and -19,

0
b C
-0.3p
Fig. 8. TEMs of RuO&ained thin sections. Change of internal particle morphology of
PMS-31 during the polymerization at W C .The monomer conversions were (a) 3.7%,(b) 31.4%,
and (c) 92.5%. The dark PSrich regions sandwiched the bright PMMA region as shown in (b).
1916 CHO AND LEE

a b

C d
Fig. 9. TEMs of Ru0,-stained thin sections. Change of internal particle morphology of
PMS30 during the polymerization at 800C. The monomer conversions are (a) 6.1%,(b)43.9%,
(c) 57.5%, and (d) 99.4%.The core-shell (PSshell, PMMAcore) morphology is seen in (b)and
(C).

Fig. 10. TEMs of RuOl-stained thin section of PMS24 (polymerizationtemperature, 800C).

The core-shell morphology (PSshell, PMMAcore) can be identified.
 PMMA-PS COMPOSITE LATEX 1917

the PSrich domain grew only at one side of a particle, and we called the
resulting morphology “halfmoonlike” (cf. Fig. 1).The latex particles in P M S
31 or PMS5, -6,and -19 grew into either the sandwichlike or halfmoonlike
shape in somewhat mixed modes. However, the major morphology of the
particles were distinct enough to classify them either as sandwichlike or
as halfmoonlike.
Interfacial tensiong could also play important roles in determining the
morphology of monomer-swollen particles, when they were dispersed in
water. Those relevant interfacial tension values were shown in Table V.
Those of PS and PS(emu1sion polymerizedbstyrene to water were greater
than those of PMMA and PMMA(emu1sion polymerizedhtyrene. However,
the interfacial tension of PS(emu1sion polymerizedbstyrene was far lower
than that of PS(bu1k polymerizedt-styrene. It indicated that the difference
between interfacial tension of PS-styrene and PMMA-styrene might not
be observable in the presence of emulsifier or the polymer with ionic ter-
minal groups.
The rate of decomposition of K2S208 was temperature-dependent, and the
concentration of 4 0 s group incorporated into the particles could be dif-
ferent under different polymerization conditions. As shown in Figure 11,
the polymerizations were virtually completed by peak time. If all the sulfate
anion radicals had been produced before peak time and successfully initiated
the styrene polymerization in the seed particles, the amounts of sulfate
groups coupled with PS molecules could be estimated, and the results are
summarized in Table VI.
As shown in Table VI, the particle morphology changed from an inverted
core-shell (PS-core, PMMA-shell) to a core-shell with the increase in the
concentration of the 4 0 group ~ coupled with PS. From these results it
could be stated that the anchoring effect exhibited by terminal --SOT group
could lead to exhibited important consequence upon the control of particle
morphology. Matsumoto and his co-workers6concluded in their work on the
poly(ethy1 acrylate) (PEAjPS system that the morphology was influenced
not by an anchoring effect but by the viscosity of reaction loci. However,
in those systems the ratios of PEA-PS were either below or equal to 1.0 so
that during the polymerization the viscosity of the polymerization loci was
greater than that of PMMA-PS and could affect to a greater degree the
morphology of the particle. Lee and Ishikawa14also studied the inkerted
TABLE V
Interfacial Tension Values of Styrene, Polymer, and Polymer Solutions to Water at 20°C
Polymer solution Interfacial tension (dynlcm)
~ ~ ~ ~ ~ _ _ _ ~ ~

Styrene 40-438
PS 32.7’
PS + styreneb 33.2
PS + styrene’ 22.3
PMMA 26.0
PMMA + styrened 20.5
*The values were cited from Ref. 25.
Bulk polymerized PS was used (concn, 1.0 w t %).
Emulsion polymerized PS was used (concn, 1.0 wt %). From the reprecipitated PS, emulsifier
was extracted with methanol in Soxhlet extractor for 4 days.
Emulsion-polymerized PMMA was used (concn, 1.0 wt %).
1918 CHO AND LEE

Tlme
Fig. 11. Relationship between conversion and time. At the peak time the polymerization
(ca. 92%) was virtually complete.

corc+shell morphology, but in their systems the functionality of monomeric

X O O H was too excessive to be affected by the terminal functional groups
introduced by ionic initiators.
In all the PMMA-seeded polymerizations of styrene, PS- and PMMA-
styrene phases normally separated clearly before the autoacceleration stage
took place. Along with the polymerization the morphology of particles went
through transformations to the various types. In this transformation process
the mobility of polymer molecules were restricted due to the high viscosity
of polymerization loci. The polymerization and the volume contraction in
the PSrich region proceeded uniformly, but in the PMMA-rich region the
polymerizing PS molecules could not migrate into the PSrich region, re-
sulting in the formation of isolated small PS-rich domains in the PMMA-
TABLE VI
Amounts of -SOT Groups Coupled with PS Molecules per Unit Weight of PS.
Peak Polymerization
Experimental time temp Wm-b
no. (mid (“C) (x 104 g/g) Morphology
PMS54 306 80 1.14 Inverted core-shell
-19 230 60 2.04 Halfmoonlike
31 588 60 4.97 Sandwichlike
-52 480 60 16.5 Core-shell
-24 97 80 19.6 Core-shell
-30 135 80 24.8 Core-shell
-40 31 80 22.3 Raspberrylike
*The decompasition rate constants of K$3,08 were 3.16 x and 9.16 X 8-l at W C
and W C ,respectively (cf. Ref. 26).
Weight of So,-coupled with unit weight of PS.
 PMMA-PS COMPOSITE LATEX 1919

rich region, and volume contraction in the PMMA-rich region could not
take place in an orderly manner. Thus, the PMMA-rich region of the com-
pletely polymerized particles became mixed with small PS domains [Figs.
qc), 8(c), 9(d), and 101 forming various anomalous particle shapes [Figs. l(a),
I@), 3(d), 4(d), 5(c), 12(b), and 12(c)].
In the cases of PMS24, -30,and -53 the formation of the voidcontaining
particles in small conversion level were confirmed by TEMs [Fig. 4(b)] and
SEMs [Fig. 12(a)].These morphologies have been reported in soapfree sty-
rene emulsion polymerizations,28PMMA-PS'O and PBA-PSI1 composite la-
texes. According to Cox and co-workers,28the voids in PS particles were
formed by the vaporization of monomer from the swollen particles during
the sample preparation for TEMs. The swollen particles were composed of
monomer-rich cores and polymer-rich shells. In the formation mechanism
of PBA-PS void particles," the PBA-rich region was effused out from the
PBAcore/PSshell composite particles in the process of the specimen p r e p
aration for electron microscopy because of the low glass transition tem-
perature (-54°C) of PBA. However, the formation mechanism of PMMA-
PS particles such as that are described above were not suggested.1°
A schematic representation of styrene-free core-shell, halfmoonlike, sand-
wichlike, and inverted core-shell type particles, and the corresponding par-
ticles swollen by styrene, were shown in Figure 13. The shapes of swollen
particles in a latex state are assumed from the premises that they must
maintain the phase-separated sphere morphology because of the surface
tension and their limited solubility in monomer. These sphere particles
became the void-containing or other shapes according to their initial swollen
state, being caused by initiator anchoring effects. It is obvious that the
volume contraction of PMMA-rich region is much larger than that of P S
rich region during the specimen preparation for electron microscope, con-
sequently generating the various morphologies as illustrated in Figure 13.
Effects of Viscosity
In the previous section the anomalous particle morphology at the final
stage of polymerization was discussed. Since these phenomena were as-

Fig. 12. SEMs of PMS latexes. Three-dimensional shapes of (a) PMS30-4 (monomer con-
version, 26.9%),(b) PMS3O (99.4%)and (c) PMS3l (92.5%)are shown. The voidcontaining
(a), the cor-hell (b), and the sandwichlike (c) morphologies can be deduced.
1920 CHO AND LEE
styrene- monomer-
-swollen -free

lnverted c o r e - s h e l l

h a l f moon-11 Ire
@a
sandwl ch -11Ire

core-shell

Fig. 13. Schematic representation of various morphologies of styrene-swollen and monomer-

PMMA-rich region; (e)PS-rich region.
free state particles: (0)

sumed to be caused in part by the high viscosity of polymerization loci, the

effects of viscosity on the particle morphology were examined.
Effects of High Viscosity. In the semibatch or the batch (swelling ratio
I 1.0) seeded emulsion polymerizations, the polymerization loci could be
kept highly viscous (Table VII). The particle morphology of PMMA-seeded
emulsion polymerization in highly viscous systems (semibatch and batch)
was studied by soap titration methods by Okubo and his co-w~rkers.~ Ac-
cording to their work, the morphology of the latex particles obtained in
semibatch polymerization changed from the initial surface localization of
PS to polymeric oil-in-oil (PO01 dispersion of PS in PMMA matrix and then
to raspberrylike types. Above 60% of total PS content the particles finally
became spherical again. In batch polymerizations the morphology of com-
posite particles changed in the similar way to that of semibatch polymer-
ization except the initial stage in which the surface localization of PS on
a PMMA particle took place. According to these authors, PS molecules
formed on the surface of PMMA particles were crushed into the interior
of a particle (PO0 dispersion stage) by more hydrophilic PMMA to form
the raspberrylike particles.
During the semibatch polymerization of PMS-45 the apparent particle
shape changed from the spherical, the raspberrylike to the irregular shapes.
TABLE VII
Polymerizations under High Viscosity Conditions
Experimental no. PMS36 PMS37 PMS-40 PMS-45 PMS51
Ingredients'
Water 95 95 95 95 95
PMMA (species) 5 (-5) 5 (-5) 5 (-4) 5 (-6) 5 (-7)
Styrene 15 15 5 15 5
KzSzOLl 0.1 0.1 0.1 0.1 0.1
KOH (0.W 5 5 5 5 5
Polymerization
method Semibatch Semibatch Batch Semibatch Batch
Dropping rate
(mL/min) 0.6 0.6 0.6
Polymerization
temp (OC) 80 60 80 60 80
Quantities are in weight parts.
 PMMA-PS COMPOSITE LATEX 1921

And the monodispersed particle size of seed latex particles became gradually
broadened as PS content increased. These nonuniform shapes appeared
regardless of the polymerization temperature [Figs. 14(a), 14(b)]. We at-
tributed this phenomenon to the fact that the monomer distribution
throughout the particles was not uniform because of the insufficient mixing
caused by the high viscosity and the monomer in the monomer-starved
particles was polymerized rapidly by Tromsdorff effect. The particle mor-
phology was certainly affected by high viscosity as pointed out by Okubo
and his co-w~rkers.~ However, we found that the anchoring effect of
-SOT groups contributed significantly to the localization of PS-rich do-
mains at the surface area of the particles. And as shown in Figure 14(c),
the PMMA-rich region was surrounded by post-polymerized PS (PS/PMMA
= 3). The smooth surfaces of PS-rich region in Figures 14(a) and 14(b)were
formed by the increased PS volume spreading locally over the particle.
The changes in particle morphology taking place in the batch polymer-
ization (styrene/PMMA = 1)are shown in Figure 15. At the initial stage
of polymerization (Fig. 15(a); conversion, 5.5%), the PS phase was not dis-
tinguishable. But at the level of 28% conversion the PS-rich domains were
identifiable at the surface region of particles, even though the contrast was
not clear. Toward the end of polymerization the clear PS-rich domains
surrounded almost all the surface of PMMA-rich region. Although in these
polymerizations the monomer concentration in the particle was greater
than in semibatch polymerizations, the mobility of polymer molecules were
also limited by the high viscosity. Accordingly, the mobility of the PS-rich
region was restricted, and PS domains were formed at the surface area of
particles.
Effects of Low Viscosity. The viscosity of polymerization loci could be
lowered by the control of polymer molecular weight or by the addition of
an inert solvent (Table VIII).
When the particle morphology of PMS-32 (Fig. 16) was compared with
that of PMS-31 [Fig. 8(c)], PS-rich domains in PMMA-rich phase were re-

Fig. 14. Latex TEMs (a,b)and "EM of Ru04-stained thin section (c). (a)PMS36 [polymer-
ization temp, 8 W q ; @,c) PMS-37 [polymerization temp, 60°C; @), (c)] gave almost the same
morphologies. In (c) the PMMA-rich region was surrounded by PSrich region (core-ehell
morphology).
1922 CHO AND LEE

Fig. 15. TEMa of Ru04-stained thin sections. Change of internal particle morphology of
PMS-51 at 8OT (styrene/PMMA= 1.0) The monomer conversions were (a) 5.5%, (b) 27.0% and
(c) 99.2%. The dark PS domains gradually covered and were clearly identifiable over the
particle surface.

duced, and the easier phase separation of the PSrich region from the
PMMA-rich region was observed. The interparticle coalescence of PSrich
phases was also observed in the former case (PMS32). These results could
be attributed to the better mobility of PS molecules of low molecular weight.
Swelling of the polymerizing particles with inert solvents was chosen by
many authors for kinetic studiesB and for particle size enlargemenLm For
the purpose of decreasing the viscosity in polymerization loci, toluene was
added into the polymerization systems PMS-41 and -42. The rate of poly-
merization decreased, and the autoacceleration stages were not observed.
These results were based on the viscosity reduction of the polymerization
loci by dilution of the system with toluene. Consequently, polymer molecules
became more mobile and facilitated the phase separation of PSrich and
PMMA-rich phases, and the low concentration of PS in PMMA-rich regions
provided the smooth particle surface (Fig. 17).
The high viscosity at the autoacceleration stages restricted the long-range
movements of polymer molecules, thus making the merge of PSrich do-
mains in the PMMA-rich region difficult. This restricted mobility of polymer
molecules could be relieved by the addition of cosolvent, and the particle
TABLE VIII
Polymerization under Low Viscosity Conditions"
~~~ ~

PMS-32 PMS41 PMS42

Ingredients
Water 95 95 95
PMMA (seed) 5 (PMMA4) 5 (-5) 5 (-5)
Styrene 15 15 15
KOH (0.1M 5 5 5
K&O, 0.1 0.1 0.1
Toluene 5 5
CHBr, 0.1
Conditions
Polymerization temp ("C) 60 80 60
8 Quantities are in weight parts.
 PMMA-PS COMPOSITE LATEX 1923

Fig. 16. TEM of RuO,-stained thin section of PMS-32 at 60°C.Because of the low molecular
weight of PS, when polymerized with CHBr,, the phase separation between PS-rich region
and PMMA-rich domains and the interparticle coalescence occurred easily.

morphology was transformed to the thermodynamically stable form. The

representative morphology of thus obtained particles is shown in Figure
18. The raspberry PMS-40 particles were transformed to the halfmoonlike
type by the merge of PS domains.
According to the calculated value W , (cf. Table VI) the equilibrium
swelling morphology of PMS-40 should be the core-shell type, but it was
the halfmoonlike. Because of the high viscosity of polymerization loci, the
polymerization was autoaccelerated, and almost all the monomer was poly-
merized not by incoming initiator radicals but by unterminated radicals.
Therefore, the fraction of sulfate anion radicals which were coupled with
PS exhibiting anchoring effect was smaller than the totid decomposed sul-
fate anion radicals, and the halfmoonlike morphology was formed.

Fig. 17. TEM of RuO,-stained thin section of PMS-42 at 60°C.Because of low Viscosity in
the toluene-swollen particle the phase separation between PSrich and PMMA-rich regions
occurred clearly.
1924 CHO AND LEE

C d
Fig. 18. Latex TEMs (a,b)and TEMs of Ru04-stainedthin sections (c,d)of PMS-40. Before
benzene swelling the particles were shown raspberry morphology (a,c). But after benzene
swelling the particle morphology was changed to the halfmoonlike morphology (b,d).

CONCLUSION
The morphology of composite latex particles prepared by PMMA-seeded
emulsion polymerization of styrene could be controlled by various poly-
merization reaction parameters. In batch polymerizations, the equilibrium
swelling at each polymerization temperature could be attained during the
warmup time (about 1h). Thus the long swelling time at room temperature
before polymerization did not affect the particle morphology. The decom-
position rate of K2S206 remained unchanged in the pH range 3-12, and in
this pH range the particle morphology was not affected. The anchoring
effect exhibited by the terminal functional groups originated from the ionic
initiators affected significantly the particle morphology. Oil-soluble hydro-
phobic initiator AIBN and water-soluble but less hydrophilic initiator
ABCVA gave inverted core-shell morphology. When water-soluble hydro-
philic initiator K2S206 was employed, the particle morphology changed from
the inverted core-shell, the halfmoonlike, the sandwichlike to the core-
shell type. These morphological variations were dictated by the amount of
the bounded -SOT on PS phase, and the latter was controlled by the
 PMMA-PS COMPOSITE LATEX 1925
TABLE IX
Morphology Change of PMMA-PS Composite Particles Observed under Various
Polymerization Conditions.
Lar viscosity

Conversicm increase

High Vlrcosity
IPS/- ratio - increase 1

a (0)
PMMA-rich region; (@) PSrich region.

initiator concentration and the polymerization temperature. Viscosity of

polymerization loci controlled the mobility of polymer molecules. Therefore,
the high viscosity of polymerization loci gave the uneven morphology of
particles in which the PS-rich domains were present in the PMMA-rich
region. However, at low viscosity the smooth particle and the clear phase
separation between the PS-rich and PMMA-rich regions resulted. In the
semibatch polymerizations the monodispersity in particle size were not
obtained when monomer conversion increased, and, at PS/PMMA = 3.0,
the PS shelled the PMMA-core.
A schematic representation of different morphological transformations
we have observed in the present investigation is summarized in the Table
IX.

We are indebted to Mr. C. M. Oh, Mr. S. A. Song, and Miss I. S. Han of Lucky Central
Research Institute for the electron microscopy works.

References
1. E. Schmidt, U.S. Pat. 3,673,133(1972).
2. W.V. Smith, J. Am. Chem. Soc., 70,3695 (1948).
3.L. J. Hughes and G. L. Brown, J. Appl. Polym. Sci., 5 , 580 (1961).
4. T.R. Paxton, J. Colloid Interface Sci., 31, 19 (1969).
5. R. A. Dickie, M.-F. Cheung, and S. Newman, J. Appl. Polym. Sci., 17,65 (1973).
6.T.Matsumoto, M. Okubo, and S. Shibao, Kobunshi Ronbunshu, 33,575 (1976).
7. M. Okubo, Y. Katsuta, A. Yamada, and T. Matsumoto, Kobunshi Ronbunshu, 36,459
(1979).
8.M. Okubo, Y.Katsuta, and T. Matsumoto, J. Polym. Sci., Polym. Lett. Ed.,18,481(1980).
9. M. Okubo, A. Yamada, and T. Matsumoto, J. Polym. Sci., Polym. Chem. Ed., 16,3219
 CHO AND LEE

(1981).
10.M. Okubo, M. Ando, A. Yamada, Y. Katsuta, and T.Matsumoto, J. Polym. Sci., Polym.
Lett. Ed.,19,143 (1981).
11. M. Okubo, Y. Katsuta, and T. Matsumoto, J. Polym. Sci, Polym. Lett. Ed., 20,45 (1982).
12. D.I. Lee, Am. Chem. Soc., Div. Org, Cwt. P h t . Chem., Pap., 43,622(1980).
13. D.I. Lee, Preprints of the 2nd Japan-Korea Joint Symposium on Polymer Science and
Technology, Kyoto, 1980,p. 37.
14. D.J. Lee and T. Ishikawa, J. Polym. Sci, Polym. Chem. Ed.,21, 147 (1983).
15.T.I. Min, A. Klein, M. S. El-Aasser, and J. W. Vanderhoff, J. Polym. Sci, Polym. Chem.
Ed., 21,2845 (1983).
16. K. Kato, Jpn. Plast., 2(Apr.), 6 (1968).
17. (a) J. S.Trent, J. I. Scheinbeim, and P. R. Couchman, J. Polym. Sci, Polym. Lett. Ed.,
19,315 (1981);(b)J. S.Trent, J. I. Scheinbeim, and P. R. Couchman, Macmnwlecuh, 16,589
(1983).
18. L. H. Sperling, T.-W. Chiu, and D. A. Thomas, J. Appl. Polym. Sci, 17,2443(1973).
19. W. V. Smith and R. H. Ewart, J. Chem. Phys., 16,592 (1948).
20. M. Morton, S.Kaizerman, and M. W. Altier, J. Colloid Sci., 9,300 (1954).
21. Y. Chung-Li, J. W. Goodwin, and R. H. Ottewill, Ptwgr. ColL Polym. Sci, 60,163 (1976).
22. S.M. Hasan, J. Polym. Sci, Polym. Chem.20, 3031 (1982).
23. D.C. Blackley, Emulsion Polymerizatwn, Applied Science, London, 1975,pp. 155-250.
24. R. J. Kern, J. Polym. Sci., 21, 19 (1956).
25. B. R. Vijayendran, Polymer Collicls II, R. M. Fitch, Ed., Plenum, New York, 1980,pp.
71-83.
26. Polymer Handbook, J. Brandrup and E. H. Immergut, Eds., Wiley, New York, 1975,p.
11-40.
27. G. E. Molau: (a) J. Polym. Sci. Part A, 3, 1267 (1965);(b) 3, 4235 (1965).
28. R. A. Cox, M. C. Wilkinson, J. M. Creasey, A. R. Goodall, and J. Hearn, J. Polym. Sci.,
Polym. Chem. Ed.,15,2311 (1977).
29. D. C. Blackley and A. Haynes, Br. Polym. J., 9,321 (1977).
30. J. Ugelstad and P. C. Mork, Adu. Colloid Interface Sci., 13, 101 (1980).
Received June 29,1984
Accepted August 30,1984