Synthesis and characterization of organic/inorganic hybrid nanoparticles: Kinetics of surface-initiated atom transfer radical polymerization and morphology of hybrid nanoparticle ultrathin films

Article abstract of DOI:10.1021/ma034188t

The synthesis of hybrid nanoparticles was conducted by the atom transfer radical polymerization (ATRP) of styrene and (meth)acrylates from colloidal surfaces. Colloidal initiators were prepared by the functionalization of silica colloids with 2-bromoisobu

Full text of DOI:10.1021/ma034188t

5094
Macromolecules 2003, 36, 5094-5104
Synthesis and Characterization of Organic/Inorganic Hybrid
Nanoparticles: Kinetics of Surface-Initiated Atom Transfer Radical
Polymerization and Morphology of Hybrid Nanoparticle Ultrathin Films
J effr ey P yu n , Sh iju n J ia , Tom a sz Kow a lew sk i,* Ga r y D. P a tter son , a n d
Kr zysztof Ma tyja szew sk i*
Department of Chemistry, Carnegie Mellon University, Center for Macromolecular Engineering,
4400 Fifth Avenue, Pittsburgh, Pennsylvania 15213
Received February 13, 2003; Revised Manuscript Received May 9, 2003
ABSTRACT: The synthesis of hybrid nanoparticles was conducted by the atom transfer radical
polymerization (ATRP) of styrene and (meth)acrylates from colloidal surfaces. Colloidal initiators were
prepared by the functionalization of silica colloids with 2-bromoisobutyrate groups. ATRP from colloidal
surfaces was then performed to attach well-defined homopolymers and block copolymers to an inorganic
core. Kinetics of the ATRP of styrene (Sty), n-butyl acrylate (BA), and methyl methacrylate (MMA) under
identical reaction conditions were investigated. Hybrid nanoparticles containing block copolymers of pSty-
b-pBA (Mn ) 22 300; Mw/Mn ) 1.20), pMMA-b-pBA (Mn ) 29 400; Mw/Mn ) 1.28), and pBA-b-pMMA (Mn
) 17 300; Mw/Mn ) 1.28) were prepared, and hydrolysis of silica cores by hydrofluoric acid treatment
enabled characterization of cleaved copolymers using size exclusion chromatography and 1H NMR.
Ultrathin films of hybrid nanoparticles were examined using transmission electron microscopy and atomic
force microscopy.
In tr od u ction
the degree of polymerization (DP) of each tethered
segment as well as of the functionality of the selected
monomers enabled precise engineering of both colloidal
surfaces and of the properties of the resulting hybrid
nanoparticle. Ultrathin films of the hybrid nanoparticles
containing block copolymers tethered to SiO1.5 colloids
were also reported to yield phase-separated microstruc-
tures when cast onto mica.41 Of particular interest was
the effect of composition, blocking sequence, and DPn
of each tethered polymeric segment on the organization
of hybrid nanoparticle (sub)monolayers deposited onto
surfaces.
Herein, we report the synthesis and characterization
of hybrid nanoparticles composed of a silica colloidal core
and of an outer shell of tethered homopolymer/block
copolymers. In contrast to our previous reports, com-
mercially available silica nanoparticles were employed
as colloidal initiators, which greatly facilitated scale-
up synthesis. Kinetics of surface-initiated ATRP of
styrene and (meth)acrylate monomers was investigated
under identical reaction conditions and compared with
ATRP using monomeric initiators. Various core-shell
colloids containing tethered AB diblock copolymers were
synthesized using different combinations of styrene and
(meth)acrylates in ATRP reactions from surfaces.
The synthesis of nanocomposite materials from (co)-
polymers and colloids has been widely investigated. The
preparation of copolymers possessing incompatible seg-
ments yields materials ordered on the nanoscale via
microphase separation in the solid state1 or self-as-
sembly in solution.2 Colloids of uniform size and precise
morphology have been synthesized from a variety of
emulsion, precipitation, and “sol-gel” approaches.3-5
The combination of these two materials enables the
preparation of a hybrid material composed of a well-
defined colloidal platform and copolymers capable of
organization into distinct domains.6,7
The preparation of organic/inorganic hybrid materi-
als8,9 composed of organic (co)polymers and inorganic
colloids has been recently pursued as a route to combine
the advantageous properties of both classes of macro-
molecules into one material. In these methodologies, the
use of living or controlled polymerization10 techniques
has been critical to incorporate organic (co)polymers of
precise molar mass, composition, and functionality to
the inorganic substrate. Recently, controlled/living radi-
cal polymerization, namely, atom transfer radical po-
lymerization (ATRP),11-19 was conducted from colloidal
surfaces to prepare hybrid nanoparticles composed of
an inorganic colloidal core and chain end immobilized
organic (co)polymers.20 The functionalization of various
colloids, including silica (SiO2),21-27 gold,28-30 silver,31
germanium,32 PbS,33 carbon black,34 iron oxides,35 metal
oxide systems,36-38 and other surfaces,39,40 was achieved,
allowing for subsequent application as initiators for the
ATRP of styrene and (meth)acrylate monomers.
Exp er im en ta l Section
Ma ter ia ls. n-Butyl acrylate (99%, Acros), styrene (99%,
Aldrich), and methyl methacrylate (99%, Acros) were stirred
over calcium hydride (coarse granules, 95%, Aldrich) overnight
and distilled before use. 1-(Chlorodimethylsilyl)propyl 2-bro-
moisobutyrate was prepared using previously reported proce-
dures.42 Copper(I) bromide (99.999%, Aldrich) and 4,4′-(di-5-
nonyl)-2,2′-bipyridine (dNbpy) were purified and prepared
according to previously reported procedures.43 Copper(II)
bromide (99.999%, Aldrich), Aliquat 336 (Aldrich), toluene
(Fisher), hydrofluoric acid (50 vol % HF, Acros), and colloidal
silica (30 wt % silica in methyl isobutyl ketone dispersion,
effective diameter ) 20 nm, MIBK-ST, Nissan) were used as
received.
Previously, our group reported the preparation of
hybrid nanoparticles with block copolymers covalently
attached to a polysilsesquioxane (SiO1.5
methodology was a versatile route to prepare multilay-
ered core-shell colloids by the sequential surface-
initiated ATRP of styrenes and acrylates. Control of both
41
)
core. This
10.1021/ma034188t CCC: $25.00 © 2003 American Chemical Society
Published on Web 06/20/2003

Macromolecules, Vol. 36, No. 14, 2003
Organic/Inorganic Hybrid Nanoparticles 5095
Ch a r a cter iza tion . Size exclusion chromatography (SEC)
was performed in tetrahydrofuran at 35 °C using a Waters
510 pump set to a flow rate of 1 mL/min, three Styragel
columns (Polymer Standards Service, pore sizes 105, 103, and
102 Å), and a Waters 2410 refractive index detector. Calcula-
tions of molar mass were determined using the PSS software
using a calibration based on linear polystyrene and poly-
(methyl methacrylate) standards (from PSS). 1H NMR analysis
was done on a 300 MHz Bruker spectrometer using the
Tecmag software. Elemental analysis was conducted by Mid-
west Microlab (Indianapolis, IN). Tapping-mode atomic force
microscopy (AFM) analysis was carried out using the Nano-
scope-III Multimode system (Digital Instruments, Santa Bar-
bara, CA). The images were acquired in air with standard
silicon TESP probes (nominal spring constant and resonance
frequency respectively 50 N/m and 300 kHz). In the case of
deformable hybrid nanoparticle monolayers, rigid colloidal
silica cores were well contrasted from the deformable polymer
matrix with the increase of the effective tapping force, adjusted
by decreasing the set-point ratio A/A0, where A0 and A denote
respectively the unperturbed cantilever amplitude and “tap-
ping” amplitude.3 Samples for AFM were prepared by spin-
coating hybrid particle solutions (1 mg/mL chloroform solution,
rotational speed 4000 rpm) onto freshly cleaved mica. Trans-
mission electron microscopy was conducted using a Hitachi
H-7100 electron microscope. TEM samples of nanoparticles
and hybrid nanoparticles were prepared by casting one drop
of a dilute colloid solution onto a carbon-coated copper grid.
The light scattering setup consisted of a Spectra Physics model
2020 argon ion laser operated at 514.5 nm. Scattered light was
measured at 90° with a Brookhaven Instruments model BI-
240 goniometer. Intensity measurements and correlation
function calculations were carried out with a Brookhaven
Instruments model BI-9000 correlator. Particle sizes were
determined using the method of cumulants.
Syn th esis of 2-Br om oisobu tyr a te F u n ction a l Silica
Colloid s. The silica dispersion (52 g of 30 wt % SiO2 in methyl
isobutyl ketone (MIBK)) was added to a 100 mL three-neck
flask with magnetic stir bar and fitted with a reflux condenser
and gas inlet adapter. 1-(Chlorodimethylsilyl)propyl 2-bro-
moisobutyrate (6.0 mL, 27 mmol) was then added via syringe,
and the orange, slightly opaque dispersion became cloudy. The
reaction mixture was gently refluxed for a period of 24 h and
cooled to room temperature. Hexamethyldisilazane (HMDZ)
(6.0 mL, 28 mmol) was then added to the dispersion, stirred
for 3 h at room temperature, and then slightly heated for 6-8
h. A white solid precipitated from the dispersion after reflux
of HMDZ. The precipitate was removed by centrifugation, and
the clear orange dispersion was then added dropwise into
methanol/H2O (4:1 vol) to precipitate colloids. After recovery
by filtration, particles were redissolved in 50 mL of tetrahy-
drofuran (THF) and precipitated into hexane (500 mL).
Particles were allowed to settle in the hexane mixture for a
few hours, and approximately 400 mL of the top clear hexane
fraction was decanted. Fresh hexane (400 mL) was added to
the remaining cloudy colloid mixture, and the mixture was
stirred for 30 min. Particles were again allowed to settle for a
few hours, and the hexane layer was decanted. These washing
cycles were repeated 12 times. Particles were then recovered
by centrifugation, and the wet gel was dried under vacuum
(1-5 mm Hg) overnight. A slightly yellow granular powder
was collected (12 g, 76% yield relative to original SiO2 mass).
ATRP of Sty fr om 2-Br om oisobu tyr a te F u n ction a l
Colloid s. Silica colloidal initiators (500.0 mg, 0.13 mmol), Cu-
(I)Br (12.0 mg, 0.13 mmol), Cu(II)Br2 (2.0 mg, 0.01 mmol), and
dNbpy (121 mg, 0.29 mmol) were added to a 25 mL Schlenk
flask containing a magnetic stir bar. The flask was fitted with
a rubber septum and evacuated (1-5 mm Hg) for a period of
5 h. The flask was then backfilled with nitrogen, and the flask
was evacuated again for 5 min, followed by additional back-
filling with nitrogen. This evacuation/backfilling cycle was
repeated, and then Sty (6.9 g, 67 mmol) (bubbled for 1 h with
nitrogen before use) was added to the flask via syringe. The
rubber septum fitted on the Schlenk flask was replaced with
a greased glass stopper under high nitrogen purge, and the
reaction mixture was homogenized by agitation on a vortex
mixer for 5-10 min. Reaction flasks were then placed in a 90
°C oil bath. Samples were taken periodically via syringe for
kinetic analysis of the polymerization. After 20.2 h, a monomer
conversion of 14% was reached, as determined from gravimet-
ric analysis. Samples were diluted with tetrahydrofuran,
filtered through neutral alumina, concentrated in vacuo to a
volume of approximately 10 mL, and precipitated into 500 mL
of methanol, yielding a white solid (Mn
Mw/Mn ) 1.21).
) 9200;
SEC cleaved pSty
ATR P of BA fr om 2-Br om oisob u t yr a t e F u n ct ion a l
Colloid s. Silica colloidal initiators (500.0 mg, 0.13 mmol), Cu-
(I)Br (12.0 mg, 0.130 mmol), Cu(II)Br2 (2.0 mg, 0.010 mmol),
and dNbpy (121.0 mg, 0.29 mmol) were added to a 25 mL
Schlenk flask containing a magnetic stir bar. The flask was
fitted with a rubber septum, and the flask was evacuated (1-5
mmHg) for a period of 5 h. The flask was then backfilled with
nitrogen, and the flask was evacuated again for 5 min, followed
by additional backfilling with nitrogen. This evacuation/
backfilling cycle was repeated, and then BA (8.6 g, 67 mmol)
(bubbled for 1 h with nitrogen before use) was added to the
flask via syringe. The rubber septum fitted on the Schlenk
flask was replaced with a greased glass stopper under high
nitrogen purge, and the reaction mixture was homogenized
by agitation on a vortex mixer for 5-10 min. The reaction
flasks were then placed in a 90 °C oil bath. Samples were taken
periodically via syringe for kinetic analysis of the polymeri-
zation. After 52.6 h, a monomer conversion of 10% was
reached, as determined from gravimetric analysis. Samples
were diluted with tetrahydrofuran, filtered through neutral
alumina, concentrated in vacuo to a volume of approximately
10 mL, and precipitated into solution of methanol (400 mL)
and deionized water (50 mL), yielding a clear viscous oil
(Mn
) 6800; Mw/Mn ) 1.26).
SEC cleaved pBA
ATRP of MMA fr om 2-Br om oisobu tyr a te F u n ction a l
Colloid s. Silica colloidal initiators (500.0 mg, 0.13 mmol) were
added to a 25 mL Schlenk flask containing a magnetic stir
bar. The flask was then fitted with a rubber septum and
evacuated (1-5 mmHg) for 1 h, followed by the addition of
MMA (6.5 g, 65 mmol) (bubbled for 1 h with nitrogen before
use). To a separate 4 mL vial containing a magnetic stir bar
were added Cu(II)Br2 (2.0 mg, 0.013 mmol), dNbpy (10.0 mg,
0.026 mmol), and (1.0 mL, 9.3 mmol) of MMA. The vial was
fitted with a snap-top lid, and the solids were allowed to
dissolve. The maroon solution was then transferred to the 25
mL Schlenk flask via syringe, and the entire mixture was
subjected to two freeze-pump-thaw cycles. While in the
frozen state, the rubber septum was removed and Cu(I)Br (12.0
mg, 0.13 mmol) and dNbpy (110.0 mg, 0.27 mmol) were added.
The flask was then subjected to two additional freeze-pump-
thaw cycles and was then placed in a 90 °C oil bath. Samples
were taken periodically via syringe for kinetic analysis of the
polymerization. After 10 min, a monomer conversion of 13%
was reached, as determined from gravimetric analysis. Samples
were then diluted with tetrahydrofuran, passed through
neutral alumina, concentrated in vacuo to a volume of ap-
proximately 10 mL, and precipitated in 500 mL of methanol,
yielding a white solid (Mn
1.34).
) 14 900; Mw/Mn )
SEC cleaved pMMA
Syn th esis of Hybr id Na n op a r ticles P ossessin g Block
Cop olym er s. Hybrid nanoparticles possessing grafted ho-
mopolymers of pS, pBA, and pMMA were prepared using the
same general procedure as described previously, except with
a higher ratio of monomer to initiating sites (DPn ) ∆[M]/[I]0
) 1000).
SiO2-g-(p Sty-b-p BA). SiO2-g-pSty hybrid nanoparticles
(500.0 mg, 0.029 mmol) were added to a 25 mL Schlenk flask
containing a magnetic stir bar. The flask was then fitted with
a rubber septum and evacuated (1-5 mmHg) for 1 h, followed
by the addition of BA (3.7 g, 29 mmol) (bubbled for 1 h with
nitrogen before use) under a nitrogen atmosphere. After
dissolution of the solids, two freeze-pump-thaw cycles were
performed. Cu(I)Br (4.0 mg, 0.029 mmol), Cu(II)Br2 (0.6 mg,
0.003 mmol), and dNbpy (26.0 mg, 0.063 mmol) was added to
the Schlenk while the reaction mixture was in the frozen state.

5096 Pyun et al.
Macromolecules, Vol. 36, No. 14, 2003
Two additional freeze-pump-thaw cycles were performed,
and the reaction vessel was then placed in a 90 °C oil bath.
Samples were taken periodically via syringe for kinetic
analysis of the polymerization. After 21.7 h, a monomer
conversion of 3% was reached, as determined from gravimetric
analysis. Samples were then diluted with tetrahydrofuran,
passes through neutral alumina, concentrated in vacuo to a
volume of approximately 5 mL, and precipitated in 300 mL of
Sch em e 1. Syn th etic Meth od ology To P r ep a r e Hybr id
Na n op a r ticlesa
methanol, yielding a white rubbery solid (Mn
) 22 300; Mw/Mn ) 1.20).
SEC cleaved pSty-b-pBA
a
ATRP of styrene, n-butyl acrylate, or methyl methacrylate
was conducted from silica colloidal initiators (X ) 2-bro-
moisobutyrate groups) yielding hybrid nanoparticles possess-
ing tethered homopolymers. Chain extension of these hybrid
nanoparticles resulted tethering of AB diblock copolymers in
formation of core-shell colloids with outer and inner shells of
different copolymer segments.
SiO2-g-(p MMA-b-p BA). SiO2-g-pMMA hybrid nanopar-
ticles (500.0 mg, 0.036 mmol) were added to a 25 mL Schlenk
flask containing a magnetic stir bar. The flask was then fitted
with a rubber septum and evacuated (1-5 mmHg) for 1 h,
followed by the addition of BA (4.6 g, 36 mmol) (bubbled for 1
h with nitrogen before use) under a nitrogen atmosphere. After
dissolution of the solids, two freeze-pump-thaw cycles were
performed. Cu(I)Br (10.3 mg, 0.072 mmol), Cu(II)Br2 (1.6 mg,
0.007 mmol), and dNbpy (64.0 mg, 0.158 mmol) were added
to the Schlenk while the reaction mixture was in the frozen
state. Two additional freeze-pump-thaw cycles were per-
formed, and the reaction vessel was then placed in a 90 °C oil
bath. Samples were taken periodically via syringe for kinetic
analysis of the polymerization. After 14.3 h, a monomer
conversion of 16% was reached, as determined from gravimet-
ric analysis. Samples were then diluted with tetrahydrofuran,
passed through neutral alumina, concentrated in vacuo to a
volume of approximately 10 mL, and precipitated in a solution
of methanol (250 mL) and deionized water (50 mL), yielding a
of sites per particle was determined using the following
relation:
moles of Br
g of nanoparticle
no. of sites per nanoparticle )
×
av MWnanoparticle
moles of Br
)
1 mol of nanoparticle moles of nanoparticle
The number of tethered chains was determined from gravi-
metric analysis, where the mass of incorporated polymer
relative to silica was first measured, followed by SEC of
cleaved polymer chains to determine Mn. The number of
tethered chains was then calculated as follows:
white rubbery solid (Mn
) 1.28).
) 29 400; Mw/Mn
SEC cleaved pMMA-b-pBA
no. of tethered chains per nanoparticle )
moles of grafted polymer
SiO2-g-(p BA-b-p MMA). SiO2-g-pBA hybrid nanoparticles
(500.0 mg, 0.038 mmol) were added to a 25 mL Schlenk flask
containing a magnetic stir bar. The flask was then fitted with
a rubber septum and evacuated (1-5 mmHg) for 1 h, followed
by the addition of MMA (3.8 g, 38 mmol) (bubbled for 1 h with
nitrogen before use) under a nitrogen atmosphere. After
dissolution of the solids, two freeze-pump-thaw cycles were
performed. Cu(I)Cl (3.7 mg, 0.038 mmol), Cu(II)Cl2 (0.5 mg,
0.003 mmol), and dNbpy (34.0 mg, 0.084 mmol) were added
to the Schlenk while the reaction mixture was in the frozen
state. Two additional freeze-pump-thaw cycles were per-
formed, and the reaction vessel was then placed in a 90 °C oil
bath. Samples were taken periodically via syringe for kinetic
analysis of the polymerization. After 45 min, a monomer
conversion of 3.5% was reached, as determined from gravi-
metric analysis. Samples were then diluted with tetrahydro-
furan, passed through neutral alumina, concentrated in vacuo
to a volume of approximately 5 mL, and precipitated in a
solution of methanol (250 mL) and deionized water (50 mL),
moles of nanoparticle
where
moles of grafted polymer )
mass of grafted polymer (gravimetry)
Mn,SEC cleaved chains
and
mass of silica colloid
moles of nanoparticle )
av MWnanoparticle
Resu lts a n d Discu ssion
The methodology used to prepare hybrid nanopar-
ticles is presented in Scheme 1. In the first stage of the
reaction, silica colloids possessing silanol groups on the
surface were functionalized with ATRP initiating groups.
These colloids were isolated and then redispersed in an
ATRP reaction mixture to polymerize styrene (Sty),
n-butyl acrylate (BA), or methyl methacrylate (MMA)
from the nanoparticle surface. From this approach,
hybrid nanoparticles possessing a shell of tethered
organic homopolymers and a core of silica were pre-
pared. In these hybrid nanoparticles, sufficient retention
of active end groups enabled the synthesis of tethered
AB diblock copolymers by chain extension reactions with
a different monomer.
Syn th esis of Silica Colloid a l In itia tor s for ATRP .
The colloidal initiator was prepared by the silylation of
silica nanoparticles using both a functional chlorosilane
(1-(chlorodimethylsilyl)propyl 2-bromoisobutyrate) and
hexamethyldisilazane. Colloids were preformed and
dispersed into methyl isobutyl ketone before use, allow-
ing direct addition of functional silanes to the colloidal
dispersion. Dynamic light scattering (DLS) measure-
ments of dilute solutions of the silica colloids revealed
an average effective diameter (Deff ) 2Rh) of 20 nm. The
yielding a white rubbery solid (Mn
17 300; Mw/Mn ) 1.28).
)
SEC cleaved pBA-b-pMMA
Gen er a l P r oced u r e for Clea va ge of (Co)p olym er s fr om
P a r ticles. SiO2-g-pS hybrid nanoparticles (100.0 mg) were
dissolved in 1 mL of toluene, along with Aliquat 336 (phase
transfer catalyst, 10.0 mg, 0.024). A 49% HF(aq) solution (1
mL) was added to the particle solution, and the reaction was
allowed to stir at room temperature overnight. PS was
recovered by precipitation into methanol and vacuum filtration
through fine grain glass frit, yielding a white powder (50 mg).
Similar procedures were followed for the cleavage of pMMA,
pBA, and block copolymers. Recovery of cleaved pBA was
performed after stirring in the toluene/HF mixture by evapo-
ration of volatiles in air.
Ca lcu la tion of th e Nu m ber of In itia tion Sites a n d
Teth er ed Ch a in s p er Na n op a r ticle. The number of possible
initiation sites for silica nanoparticle was determined from
elemental analysis, which yielded the total moles of Br per
gram of nanoparticle. The average molar mass of nanoparticles
was determined by first measuring the average hydrodynamic
radius (Rh) of bare silica nanoparticles using dynamic light
scattering. Assuming a spherical geometry for silica colloids
(V ) 4/3πr3) and a density Fsilica ) 2.07 g/cm3,44 the average
molar mass of nanoparticle was calculated. The total number

Macromolecules, Vol. 36, No. 14, 2003
Organic/Inorganic Hybrid Nanoparticles 5097
Ta ble 1. Con d ition s for th e ATRP of Sty, BA, a n d MMA fr om 2-Br om oisobu tyr a te F u n ction a l Na n op a r ticles
reaction
time (min)
experiment
colloidal initiator
monomer
BA
conditionsa
convc (%)
10
Mn
Mw/Mn
SEC
1
2-bromoisobutyrate
functional silica
2-bromoisobutyrate
functional silicab
2-bromoisobutyrate
functional silica
2-bromoisobutyrate
functional silica
2-bromoisobutyrate
functional silica
500:1:1:0.1:2.2, 90 °C
3156
6 800
9 200
1.26
2
3
4
5
6
7
Sty
500:1:1:0.1:2.2, 90 °C
500:1:1:0.1:2.2, 90 °C
1000:1:1:0.1:2.2, 90 °C
1000:1:1:0.1:2.2, 90 °C
1000:1:1:0.1:2.2, 90 °C
2000:1:1:0.1:2.2, 90 °C
1211
10
14
13
13
9
1.21
1.34
1.21
1.41
1.28
1.23
MMA
sty
14 900
18 200
17 600
12 000
30 100
1872
17
MMA
BA
2-bromoisobutyrate
functional silica
2-bromoisobutyrate
functional silica
3690
37
7
MMA
11
8
9
SiO2-g-pSty175
SiO2-g-pMMA175
SiO2-g-pBA94
BA
BA
MMA
MMA
1000:1:1:0.1:2.2, 90 °C
1000:1:1:0.1:2.2, 90 °C
1000:1:1:0.1:2.2, 90 °C
3450:1:3:0.3:6.6, 90 °C
1305
860
45
3
16
3.5
4
22 300
29 400
15 000
14 200
1.20
1.28
1.46
3.10
10d
11d
SiO2-g-(pBA94-b-pMMA62
)
232
a
Conditions correspond to molar ratios of monomer, 2-bromoisobutyrate groups on silica colloids, Cu(I)Br, Cu(II)Br2, and dNbpy. All
reactions were performed in bulk monomer. 2-Bromoisobutyrate silica colloids contained 0.27 mmol of Br/1 g of nanoparticle. c Conversion
values determined using gravimetric analysis. Cu(I)Cl was used instead of Cu(I)Br.
b
d
average molar mass of silica colloids (Mw ) 5.2 × 106
g/mol) was calculated using particle size data from DLS
along with assumptions that colloids were spheres (V
) 4/3πr3), and the density was comparable to bulk SiO2
(FSiO ) 2.07 g/cm3). Elemental analysis of the function-
2
alized silica colloid confirmed the incorporation of
bromine (2.48 wt %, 0.135 mmol of Br/1 g of SiO2), and
calculations indicated that approximately 1400 initia-
tion sites per colloidal particle were present (see Ex-
perimental Section for calculation).
Hom op olym er iza tion Kin etics. In the first step
toward preparing hybrid nanoparticles, the homopo-
lymerization kinetics of Sty, MMA, and BA ATRP from
silica colloidal initiators was investigated (Table 1).
Polymerizations were performed using identical sto-
ichiometric ratios and concentrations of monomer,
catalyst, and colloidal initiator. In this comparative
kinetic study, reactions were carried out in bulk mono-
mer (7-9 M) using a high molar ratio of monomer to
initiator sites per nanoparticle, where the average
concentration of bromine containing groups was ap-
proximately 2 × 10-2 M. Colloidal initiators were diluted
to approximately 5-8 wt % relative to the selected
monomer in the reaction mixture. The catalytic system
employed in these reactions consisted of copper(I)
bromide/4,4′-di(5-nonyl)-2,2′-bipyridine (Cu(I)Br/dNbpy)
and copper(II) bromide/4,4′di(5-nonyl)-2,2′-bipyridine
(Cu(II)Br2/dNbpy) complexes at concentrations approxi-
mately 2 × 10-2 and 2 × 10-3 M, respectively. Overall,
the initial molar ratios of monomer ([M]0), initiating
sites ([I]0), Cu(I)Br, Cu(II)Br2, and dNbpy were the
following: 500:1:1:0.1:2.2. Surface-initiated ATRP reac-
tions were all performed at 90 °C. Polymerizations were
stopped at monomer conversions under 15%, as higher
conversions resulted in interparticle coupling or gela-
tion.
Figure 1a shows the ATRP kinetics of Sty, BA, and
MMA from colloidal surfaces presented as semiloga-
rithmic plots of monomer consumption with time. The
rate of BA ATRP from 2-bromoisobutyrate functional
silica colloids (1, Table 1) was slowest of the three
monomers, as a conversion of 9% required a reaction
time of 3150 min (52 h, 30 min). In the case of Sty
polymerization (2, Table 1) a monomer conversion of
14% was reached in 1210 min (20 h, 10 min). Compari-
F igu r e 1. (a) Semilogarithmic plots of monomer conversion
vs time for the ATRP of BA (filled circles), Sty (filled squares),
and MMA (filled triangles), using 2-bromoisobutyrate colloidal
initiators at stoichiometric ratios of 500:1:1:0.1:2.2 for [M]:[I]:
[Cu(I)]:[Cu(II)]:[dNbpy] at 90 °C. Conditions are listed in Table
1, entries 1, 2, 3, respectively. (b) Semilogarithmic plots of
monomer conversion vs time for the ATRP of BA (open circles),
Sty (open squares), and MMA (open triangles), using ethyl
2-bromoisobutyrate at stoichiometric ratios of 500:1:1:0.1:2.2
for [M]:[I]:[Cu(I)]:[Cu(II)]:[dNbpy] at 90 °C.
son of the apparent rate constants (kapp, i.e., slopes from
semilogarithmic plots) for Sty and BA surface-initiated
ATRP showed that the rate of Sty polymerization was
4 times faster. The rate of MMA ATRP from colloidal
initiators (3, Table 1) was the fastest as gelation due to

5098 Pyun et al.
Macromolecules, Vol. 36, No. 14, 2003
Ta ble 2. Su m m a r y of Ra te a n d Equ ilibr iu m Con sta n ts
Deter m in ed fr om th e Hom op olym er iza tion of Sty, BA,
a n d MMA in th e ATRP fr om Eth yl 2-Br om oisobu tyr a tea
interparticle radical coupling reactions occurred within
15 min. The observation of gelation in ATRP from
multifunctional initiators has also been reported in the
synthesis of (hyper)branched45 and star46 copolymers.
Values of kapp for the ATRP of MMA indicated that the
rate of polymerization was 430 times greater than
obtained in the surface initiated BA ATRP reactions.
We anticipated that the rate of MMA ATRP from
colloids would be faster than those of Sty and BA.47
However, the faster rates for Sty ATRP vs BA at 90 °C
from colloidal surfaces were unexpected, since in the
previous kinetic investigations using small molecule
initiators acrylates homopolymerized faster than sty-
rene under comparable conditions.43,48 In the surface-
initiated ATRP of MMA, Sty, and BA, 10 mol %
Cu(II)Br2 (relative to Cu(I)Br) was added at the very
beginning of the reaction to ensure efficient exchange
reactions between dormant and active species. We
anticipate on the basis of previous EPR measure-
ments49,50 that 10 mol % of Cu(II)Br2 is much more than
what is generated spontaneously in the ATRP of acry-
lates and styrene due to the persistent radical effect.51
The EPR results indicated that at the molar ratios [M]0:
[RBr]0:[Cu(I)]0 ) 100:1:1 at ∼100 °C about 6% of Cu(II)
species was formed for styrene and about 2% for
acrylates. Because of much higher dilution in the
current system (500:1:1), the concentration of spontane-
ously formed Cu(II) should be much smaller. Thus, the
addition of the same large excess of Cu(II) (10%) should
suppress the rate of ATRP of acrylate stronger than that
of styrene. To corroborate this hypothesis, the rates of
polymerization for Sty, BA, and MMA using conven-
tional ATRP initiators were pursued.
monomer
kpapp, s-1
kp, M-1 s-1
Keq
ki, M-1 s-1
MMA
Sty
BA
2 × 10-4
2 × 10-6
5 × 10-7
1615
895
56780
7 × 10-7
4 × 10-8
4 × 10-10
3.7 × 103
5.5 × 103
1.2 × 103
a
Keq values obtained from ATRP using ethyl 2-bromoisobu-
tyrate initiator, Cu(I)Br/Cu(II)Br2/dNbpy catalyst at 90 °C, as-
suming constant ratio [Cu(I)]:[Cu(II)] ) 10, [RBr]0 ) 0.013-0.018
M. Initiation rate constants at 21 °C (ki) for isobutyrate radical
addition to respective monomers (ref 56). Rate constant of propa-
gation (kp) (ref 57).
Keq ) (kpapp/kp)[Cu(II)]/([RBr][Cu(I)])
(2)
Table 2 presents pertinent rate constants and equi-
librium constants for the three monomers are presented.
The ATRP Keq values at 90 °C of MMA and Sty (Keq MMA
) 7 × 10-7, Keq Sty ) 4 × 10-8) were higher with the
smallest for BA (Keq BA ) 4 × 10-10). This agrees with
the better radical stabilization effect of phenyl substit-
uents relative to carboalkoxy group and also the higher
stability of tertiary radicals over secondary.
In direct comparisons of ATRP from colloids and small
molecule initiators, the polymerization rates were gen-
erally slower from 2-bromoisobutyrate functional nano-
particles. Using eqs 1 and 2, Keq values for the ATRP of
MMA, Sty, and BA were also calculated from 2-bro-
moisobutyrate functional colloidal initiators, with the
assumption of a constant ratio of Cu(I) to Cu(II) species.
Although ATRP rates for MMA were fast for both types
of initiators, Sty and BA polymerization were noticeably
slower from colloidal surfaces. Because of the slower
rates of polymerization, calculated values of Keq from
2-bromoisobutyrate functional silica nanoparticles were
lower than those from the ethyl 2-bromoisobutyrate
A fundamental question pertaining to colloidal initia-
tor systems is whether the kinetic behavior of ATRP
reactions is comparable to untethered, small molecule-
initiated ATRP. A key difference in colloidal systems is
the presence of initiating groups at high local concen-
trations due to immobilization on nanoparticle surfaces.
This is in direct contrast to small molecule initiators,
which are homogeneously distributed throughout the
reaction media, along with the monomer and catalyst.
Thus, the ATRP of Sty, BA, and MMA was performed
using ethyl 2-bromoisobutyrate as the initiator, employ-
ing identical concentrations of reagents as for the
colloidal initiator systems.
initiator (Keq-SiO MMA ) 7 × 10-7, Keq-SiO Sty ) 1 × 10-8
,
2
Keq-SiO BA ) 7 × 10-11). It is unlikely th2at the Keq from
2
colloidal vs monomeric initiator would be drastically
different; thus, discrepancies in calculated Keq values
arise from a variation in the ratio of Cu(I) and Cu(II)
complexes in ATRP from 2-bromoisobutyrate functional
nanoparticles. While it is known from EPR measure-
ments that less than 10 mol % Cu(II) is typically formed
in conventional ATRP systems with monomeric initia-
tor,49,50 the amount of Cu(II) formed in ATRP from
colloidal initiators may be higher.
Figure 1b shows the first-order plots of monomer
consumption with time for the ATRP of Sty, BA, and
MMA using ethyl 2-bromoisobutyrate as the initiator.
The polymerization rates followed the same trends as
for ATRP reactions from colloidal initiators, as the rate
Discrepancies observed in the ATRP kinetics from
colloids may be explained by differences in the rates of
termination of surface-immobilized radicals vs carbon-
centered radicals formed in solution. If termination
events occur faster for surface tethered radicals, then
a higher Cu(II) concentration relative to small molecule
initiator systems may be generated. These intramolecu-
lar termination events can be qualitatively measured
using SEC as the formation of dead polymer chains will
also be accompanied by an increase in polydispersity
(Mw/Mn), particularly in chain extension experiments
(i.e., block copolymerization) from SiO2-g-pSty, SiO2-g-
BA, and SiO2-g-pMMA colloidal initiators. Because of
of monomer consumption scaled as follows: Rp MMA
.
Rp Sty > Rp BA. The rates of ATRP reactions initiated
with ethyl 2-bromoisobutyrate were also faster than the
corresponding polymerizations from colloidal initiators.
The kinetic data from ATRP of three monomers using
monomeric initiator in the presence of a large excess of
Cu(II) allows comparison of ATRP equilibrium constants
for the three monomers, using available free radical
propagation rate constants, kp. The slopes of semiloga-
rithmic plots, the apparent propagation rate coefficients,
kpapp, are products of kp and concentration of radicals
[P*]. Since the equilibrium constant, Keq, equals
the large difference in Keq BA and Keq SiO BA, BA systems
2
should generate the highest amount of dead chains
relative to Sty and MMA ATRP reactions. This point
will be discussed in later sections.
Plots of Mn SEC of (cleaved) pSty vs conversion were
constructed comparing the evolution of molar mass for
polymers grown from colloidal surfaces and from mon-
Keq ) ([P*][Cu(II)])/([RBr][Cu(I)])
(1)
values were calculated, assuming a near constant ratio
of two copper species (eq 2).

Macromolecules, Vol. 36, No. 14, 2003
Organic/Inorganic Hybrid Nanoparticles 5099
F igu r e 4. Plot of Mn vs conversion for the ATRP of MMA from
2-bromoisobutyrate colloidal initiators (filled squares, Mn; filled
circles, Mw/Mn) and ethyl 2-bromoisobutyrate (open squares,
Mn; open circles, Mw/Mn) using stoichiometric ratios of 500:1:
1:0.1:2.2 for [M]:[I]:[Cu(I)]:[Cu(II)]:[dNbpy] at 90 °C.
F igu r e 2. Plots of Mn and Mw/Mn vs conversion for the ATRP
of Sty from 2-bromoisobutyrate colloidal initiators (filled
squares, Mn; filled circles, Mw/Mn) and ethyl 2-bromoisobu-
tyrate (open squares, Mn; open circles, Mw/Mn) using stoichio-
metric ratios of 500:1:1:0.1:2.2 for [M]:[I]:[Cu(I)]:[Cu(II)]:
[dNbpy] at 90 °C.
from the monomeric initiator system. Polydispersity of
cleaved pBA ranged from Mw/Mn ) 1.3-1.5.
Similar features of a slow initiation process were also
obtained for the ATRP of MMA from 2-bromoisobutyrate
functional colloids (Figure 4). For polymerizations using
ethyl 2-bromoisobutyrate as the initiator, the plot of
Mn SEC vs conversion was linear, with a high initiation
efficiency (∼90%). The polydispersity of pMMA formed
from ethyl 2-bromoisobutyrate initiators was below Mw/
Mn < 1.2. However, for the ATRP of MMA from colloids
(3, Table 1), the initiation efficiencies (20%) were low
from the initial stages of the reaction, as a molar mass
of cleaved pMMA was approximately 5000 g/mol higher
than theoretical predictions. As the reaction proceeded,
the molar mass of cleaved pMMA approached theoreti-
cal values, indicating that the initiation efficiency also
increased with reaction time. Typically, in ATRP pro-
cesses with efficient initiation, comparable initiation
efficiencies are maintained throughout the polymeriza-
tion, indicative that a constant number of chains are
formed and grown with increasing conversion. In the
case of MMA ATRP from colloids, a gradual increase in
the initiation efficiency revealed that new chains were
continuously formed with longer reaction times. Despite
the large difference in Mn theoretical vs Mn SEC values,
cleaved pMMA had low polydispersity (Mw/Mn ) 1.2-
1.3). The lower efficiency of initiation from colloidal
surfaces may be explained by steric congestion of
2-bromoisobutyrate groups. Similar behavior has been
F igu r e 3. Plots of Mn and Mw/Mn vs conversion for the ATRP
of BA from 2-bromoisobutyrate colloidal initiators (filled
squares, Mn; filled circles, Mw/Mn) and ethyl 2-bromoisobu-
tyrate (open squares, Mn; open circles, Mw/Mn) using stoichio-
metric ratios of 500:1:1:0.1:2.2 for [M]:[I]:[Cu(I)]:[Cu(II)]:
[dNbpy] at 90 °C. An expanded plot showing early conversion
points of Mn vs conversion is shown in the upper right.
omeric initiators. Hybrid nanoparticles were treated
with hydrofluoric acid to hydrolyze the silica core and
cleave tethered polymers for SEC analysis. As shown
in Figure 2, the relationship between Mn SEC and
conversion is linear for both polymerizations initiated
from small molecules and colloids. For the ATRP of Sty
using ethyl 2-bromoisobutyrate as the initiator, the
initiation efficiency (∼99%) was very high, where the
initiation efficiency is defined as the ratio of Mn theoretical
([M]0/[I]0 × conversion × FWmonomer) and Mn SEC. PSty
of low polydispersity was also obtained (Mw/Mn < 1.18).
Polymerizations of Sty from 2-bromoisobutyrate-func-
tional colloids (2, Table 1) resulted in a lower initiation
efficiency (∼70%), yielding polymers with polydispersity
in the range of Mw/Mn ) 1.2-1.3.
The polymerization of BA from both silica colloids and
ethyl 2-bromoisobutyrate was also assessed from plots
of Mn SEC vs conversion (Figure 3). A progressive in-
crease of molar mass with conversion was observed in
both plots, although deviation from linearity occurred
in the early stages of BA ATRP from colloidal initiators.
Contrary to Sty ATRP using a monomeric initiator, the
ATRP of BA using ethyl 2-bromoisobutyrate proceeded
with a lower initiation efficiency (∼80%). Polymers had
low polydispersity as determined from SEC (Mw/Mn <
1.18). The ATRP of BA from 2-bromoisobutyrate colloids
exhibited features of a slow initiation process, as the
initiation efficiency gradually increased from 50% to
80% with increasing conversion. Initiation efficiency
values eventually became comparable to those obtained
32
reported by Patten et al., using 2-bromoisobutyrate-
functional silica colloids (particle size ) 75 nm) for the
ATRP of MMA. Additionally, Miller et al.52 observed
incomplete initiation in the ATRP of deuterated MMA
using 12-arm dendritic initiators. In that report, incom-
plete initiation was observed for star polymers of low
degree of polymerization (DP). However, quantitative
consumption of initiator groups occurred at higher DP
and conversion. The particular differences between
kinetic behavior of MMA in comparison to Sty and BA
are related to the ATRP equilibrium constants for
polymerization and initiating species as well as to the
rates of the addition of isobutyrate radicals to each
monomer (cf. Table 2).
The kinetics of the ATRP of MMA using 2-bro-
moisobutyrate-functional colloids was further investi-
gated with varying ratios of monomer to initiator,
keeping equimolar stoichiometry of initiators sites and
catalyst. Because of the very fast polymerization rates
in the MMA polymerization from colloids (3, Table 1),
ATRP conditions using a 2-fold (6, Table 1) and 4-fold

5100 Pyun et al.
Macromolecules, Vol. 36, No. 14, 2003
F igu r e 5. Semilogarithmic plots of monomer conversion vs
time for the ATRP of MMA using 2-bromoisobutyrate colloidal
initiators using conditions of [M]:[I]:[Cu(I)]:[Cu(II)]:[dNbpy]:
filled triangles (500:1:1:0.1:2.2, 90 °C), filled diamonds (1000:
1:1:0.1:2.2, 90 °C), inverted triangle (2000:1:1:0.1:2.2, 90 °C).
Conditions listed in Table 1, entries 3, 5, and 7, respectively.
dilution (7, Table 1) of initiator sites and catalyst were
also investigated. Thus, in these reactions, the molar
ratios of monomer to initiator sites were as follows:
500:1 (3, Table 1), 1000:1 (6, Table 1), 2000:1 (7, Table
1).
As shown in Figure 5, the rates of the polymerization
for MMA from 2-bromoisobutyrate colloidal initiators
decreased with higher dilution of initiator and catalyst.
Polymerizations were very fast for experiments 3 and
6 (Table 1) as gelation occurred within 20 min. The
ATRP of MMA from colloidal initiators was extended
to 40 min before the formation of gels using conditions
of higher dilution (7, Table 1 and Figure 5). From the
slope and intercept of lines in first-order plots of
monomer consumption with time (Figure 5, Table 2), a
small induction period was present for all three polym-
erization conditions. Comparison of the kapp for reactions
6 and 7 relative to 3 (Table 1) revealed that the rates
of polymerization decreased by a factor of 2 and 4,
respectively. Predictions of the polymerization rates
based on the rate expression of homogeneous bulk ATRP
reactions (eq 3) implied that 2- and 4-fold dilution of
initiator concentration would reduce the rate by the
same factor, since the ratio [Cu(I)]/[Cu(II)] was constant
(∼10). This also indicates that the amount of spontane-
ously formed Cu(II) was relatively small in comparison
with that initially added.
F igu r e 6. Plots of Mn (a) and Mw/Mn (b) vs conversion for the
ATRP of MMA from 2-bromoisobutyrate colloidal initiators
using conditions of [M]:[I]:[Cu(I)]:[Cu(II)]:[dNbpy]: 500:1:1:0.1:
2.2, 90 °C (filled triangles, Mn; open triangles, Mw/Mn); 1000:
1:1:0.1:2.2, 90 °C (filled diamonds, Mn; open diamonds, Mw/
Mn); 2000:1:1:0.1:2.2, 90 °C (filled inverted triangles, Mn; open
inverted triangles, Mw/Mn). Conditions listed in Table 1, entries
3, 5, and 7, respectively.
colloids.22 For both the ATRP of Sty from 2-bromopro-
pionate functional colloids and for the MMA ATRP from
2-bromoisobutyrate nanoparticles, initiation efficiencies
were low from very low monomer conversion and
increased with reaction time. This result was unex-
pected as methyl 2-bromopropionate and ethyl 2-bro-
moisobutyrate are relatively efficient initiators for the
ATRP of Sty and MMA, respectively. Thus, the crowded
steric environment on colloidal initiator surfaces may
hinder quantitative atom transfer of halogens from alkyl
halides to copper complexes, but over time these sites
may be accessible for initiation.
It is also important to note that while gelation due to
interparticle radical coupling occurred in the ATRP of
MMA from 2-bromoisobutyrate colloidal initiators at
higher conversion, SEC did not reveal bimodal distribu-
tions in cleaved pMMA from nanoparticle gels. In
general, pMMA cleaved from gelatinous polymerization
mixtures prepared for a variety of ATRP conditions were
completely soluble after HF hydrolysis of the core and
molar mass distributions were monomodal and fairly
narrow (Mw/Mn < 1.4). While coupling reactions could
yield bimodal distributions in cleaved polymers, SEC
revealed monomodal distributions indicating that only
a few tethered chains were needed to induce gelation
in ATRP reaction mixtures. Both the high number of
tethered chains per particle (∼1000) and the viscosity
of the polymerization mixture contribute to the rapid
onset of gelation despite the absence of measurable
amounts of coupled chains from SEC.
Rp ) kpKeq[M][I] × [Cu(I)]/[Cu(II)]
(3)
The plots of Mn SEC of cleaved pMMA vs conversion
(Figure 6) were linear for all three conditions (3, 6, 7;
Table 1) employed in the ATRP from 2-bromoisobutyrate
functional colloids. The Mn SEC of cleaved pMMA was
found to progressively increased as a functional of
dilution and higher target DPn. Polydispersities of
cleaved pMMA chains were all below Mw/Mn < 1.4.
Furthermore, features of a slow initiation process were
seen in all of the conditions used in the surface-initiated
ATRP of MMA. For all conditions (3, 6, 7; Table 1) for
MMA polymerization, Mn SEC of cleaved polymers formed
at very low conversion were significantly higher than
predicted. However, as for the other systems, the
initiation efficiencies gradually increased with reaction
time.
Syn th esis of Cor e-Sh ell Colloid s. The synthesis
of hybrid nanoparticles containing tethered block co-
polymers was performed by sequential ATRP reactions
from 2-bromoisobutyrate-functional silica colloids. A
Similar kinetic behavior of slow initiation was re-
ported in the ATRP of Sty and MMA from larger silica

Macromolecules, Vol. 36, No. 14, 2003
Organic/Inorganic Hybrid Nanoparticles 5101
previous study that prepared polystyrene-block-poly-
(benzyl acrylate) block copolymers tethered to a SiO1.5
colloidal core was conducted.41 In the current report, a
more thorough investigation into the synthesis of hybrid
nanoparticles tethered with AB diblock copolymers was
explored, focusing on the variation of copolymer com-
position, blocking sequence, and DP (Scheme 1). Using
this approach, core-shell colloids composed of a hard
inorganic core and a shell of copolymers containing both
rubbery and glassy segments were prepared. Rubbery
segments were incorporated into colloidal materials by
the ATRP of BA, while the ATRP of either Sty or MMA
resulted in the attachment of glassy segments to nano-
particle surfaces. Additionally, chain extension experi-
ments from SiO2-g-pSty, SiO2-g-pBA, and SiO2-g-pMMA
qualitatively revealed the amount of dead and active
chains present in ATRP reactions from colloids.
F igu r e 7. SEC of cleaved (co)polymers from SiO2-g-pSty175
and SiO2-g-(pSty175-b-pBA42) hybrid nanoparticles: cleaved pS
(dotted line, Mn ) 18 200; Mw/Mn ) 1.21) and cleaved pS-b-
pBA (Mn ) 22 300; Mw/Mn ) 1.20).
In the first block copolymer system, hybrid nanopar-
ticles of silica-graft-(polystyrene-block-poly(n-butyl acry-
late) (SiO2-g-(pSty-b-pBA)) were prepared by the chain
extension of SiO2-g-pSty with BA. In the first step of
the synthesis, SiO2-g-pSty hybrid nanoparticles were
prepared by the ATRP of Sty from colloidal initiators,
with a slightly higher ratio of monomer to initiator sites
([M]0/[I]0 ) 1000; 4, Table 1). Under these conditions, a
monomer conversion of 13% was reached in 15 h. SEC
of cleaved pSty measured molar mass and DP (Mn
)
18 200; Mw/Mn ) 1.21; DPn pS ) 175). From gravimetric
analysis of incorporated pSty, the number of tethered
chains per nanoparticle (Nt) was calculated to be ap-
proximately Nt ) 1000 (see Experimental Section for
calculation). Chain extension with BA used similar
reaction conditions as those in the synthesis of the SiO2-
g-pSty colloid (8, Table 1). Because of the higher molar
mass of the SiO2-g-pSty colloid relative to the bare silica
colloidal initiator, a higher weight percentage of hybrid
nanoparticle solid (10 wt %, relative to monomer) was
required in the chain extension reaction with BA. Using
these conditions, the incorporation of BA units to the
SiO2-g-pSty colloid was successful, requiring 18 h to
obtain a monomer conversion of 7% before the reaction
was quenched. 1H NMR analysis of SiO2-g-(pSty-b-pBA)
hybrid nanoparticles confirmed the presence of both
pSty and pBA protons. Integration of pBA methylene
protons δ ) 4.0 ppm and pSty aromatic protons at δ )
6.8-7.2 ppm enabled determination of molar composi-
tions of the tethered segments (80 mol % pSty, 20 mol
% pBA). Since the DPn of the starting tethered pSty was
predetermined, the DPn of pBA segments was calculated
F igu r e 8. SEC of cleaved (co)polymers from SiO2-g-pMMA175
and SiO2-g-(pMMA175-b-pBA79) hybrid nanoparticles: cleaved
pMMA (dotted line, Mn ) 17 600; Mw/Mn ) 1.41) and cleaved
pMMA-b-pBA (Mn ) 29 400; Mw/Mn ) 1.28).
1) were also employed in the ATRP of MMA from silica
colloids. In the reaction mixture, SiO2-g-pMMA colloids
were present in 11 wt % relative to monomer for the
chain extension with BA. Polymerizations were fast as
a monomer conversion of 10% was achieved in 17 min
before the reaction was quenched. From SEC of the
cleaved pMMA, molar mass and DPn were determined
(Mn SEC pMMA ) 17 600; Mw/Mn ) 1.41; DPn pMMA ) 175).
From gravimetric analysis, calculation of the number
of tethered pMMA chains per particles was Nt ) 600.
The lower number of chains per particle relative to Nt pS
calculated for SiO2-g-pSty was attributed to slow initia-
tion processes, as discussed previously. In the chain
extension of SiO2-g-pMMA with BA (9, Table 1), a
conversion of 16% was obtained in 860 min (14 h, 20
1
min). H NMR analysis of the product recovered after
the chain extension reaction confirmed the presence of
both pMMA and pBA protons. Composition of the
tethered copolymers (72 mol % pMMA, 27 mol % pBA)
1
in conjunction with H NMR composition data (SiO2-g-
(pSty175-b-pBA42)). SEC of the cleaved copolymers con-
firmed an increase in molar mass relative to the starting
pSty segment (Mn SEC pSty-b-pBA ) 22 300, Mw/Mn ) 1.20;
Figure 7). The comparable polydispersities of cleaved
(co)polymers before and after chain extension indicate
that a significant number of active tethered chains were
maintained. DLS in dilute tetrahydrofuran (THF) hy-
brid nanoparticle solutions showed an increase in ef-
fective diameter for SiO2-g-pSty175 (Deff ) 55 nm) and
SiO2-g-(pSty175-b-pBA42) (Deff ) 60 nm) relative to bare
colloidal silica (Deff ) 20 nm).
Similar hybrid nanoparticles possessing a glassy
inner block and a rubbery outer segment were obtained
by the synthesis of SiO2-g-(pMMA-b-pBA) colloids. In
the preparation of the SiO2-g-pMMA precursor, higher
ratios of monomer to initiator ([M]0/[I]0 ) 1000; 6, Table
1
determined from H NMR spectroscopy was performed
by integration of pBA methylene protons δ ) 4.0 ppm
and pMMA methyl protons at δ ) 3.6 ppm. The DPn of
the tethered block copolymer was also calculated on the
basis of predetermined DPn of the pMMA segment with
composition data from 1H NMR (SiO2-g-pMMA175-b-
pBA79). SEC of the cleaved copolymer also confirmed
in an increase in apparent molar mass relative to the
first pMMA segment (Mn pMMA-b-pBA ) 29 400; Mw/Mn
) 1.28, Figure 8). The polydispersities of cleaved (co)-
polymers before and after chain extension were com-
parable to the SiO2-g-pSty-b-pBA case, showing that
while some tethered chains may undergo intramolecular
termination, a sufficient number of active chains were
retained. DLS in dilute tetrahydrofuran (THF) hybrid
nanoparticle solutions confirmed an increase in effective

5102 Pyun et al.
Macromolecules, Vol. 36, No. 14, 2003
F igu r e 9. SEC of cleaved (co)polymers from SiO2-g-pBA94 and
SiO2-g-(pBA94-b-pMMA41) hybrid nanoparticles: cleaved pBA
(dotted line, Mn ) 12 000; Mw/Mn ) 1.28) and cleaved pBA-b-
pMMA (Mn ) 15 000; Mw/Mn ) 1.46).
diameter for SiO2-g-pMMA175 (Deff ) 60 nm) and SiO2-
g-(pMMA175-b-pBA79) (Deff ) 80 nm).
Alternatively, hybrid nanoparticles possessing a rub-
bery inner block and glassy outer blocks were prepared
by changing the sequence of connectivity of pBA and
pMMA bound to the silica colloid. SiO2-g-pBA colloids
were prepared by the ATRP from a 2-bromoisobutyrate-
functional colloid (6, Table 1). BA polymerizations
required long reaction times (61 h, 30 min), reaching a
monomer conversion of 7%. SEC of the cleaved pBA
chains were performed to measure molar mass and DP
(Mn SEC pBA ) 12 000; Mw/Mn ) 1.28; DPn pBA ) 94). The
number of tethered pBA chains per colloid calculated
from gravimetric analysis was comparable to those from
SiO2-g-pS systems (Nt pBA ) 1000). Block copolymers of
pBA-b-pMMA grafted to silica colloids were prepared
by chain extension of SiO2-g-pBA with MMA. In these
reactions, the halogen-exchange technique was required,
where a bromine-terminated SiO2-g-pBA colloid was
reacted with a Cu(I)Cl complex in the presence of
MMA.53 As for the linear copolymer systems, chain
extension was observed using SEC analysis of the
cleaved product (Mn SEC pBA-b-pMMA ) 15 000; Mw/Mn )
F igu r e 10. TEM images of hybrid nanoparticle cast onto
carbon-coated copper grids: (a) 2-bromoisobutyrate functional
silica colloidal initiators; (b) SiO2-g-pSty175 hybrid nanopar-
ticles (Figure 7, SEC; 4, Table 1); (c) SiO2-g-pBA94 hybrid
nanoparticles (Figure 9, SEC; 6, Table 1); (d) SiO2-g-pMMA175
(Figure 8, SEC; 5, Table 1). Black bar ) 200 nm.
cating efficient silylation of surface silanol groups on
nanoparticle surfaces had occurred. The particle size of
SiO2 colloidal initiators was polydisperse, although the
majority of nanoparticles possessed Deff e 20 nm. TEM
of hybrid nanoparticles for pSty, pBA, and pMMA
systems confirmed the assembly of ultrathin films
consisting of a matrix of tethered homopolymer and
dispersed silica cores arising from entanglements of
neighboring polymer chains.21,41 The distance between
the particles was governed by the radius of gyration of
tethered chains filling space between silica cores. The
majority of silica colloids imaged in ultrathin films of
SiO2-g-pSty175 (Figure 10b), SiO2-g-pBA94 (Figure 10c),
and SiO2-g-pMMA175 (Figure 10d) hybrid nanoparticles
possessed Deff e 20 nm, consistent with sizes observed
for bare silica colloid samples.
Hybrid nanoparticles tethered with block copolymers
were cast as ultrathin films onto carbon-coated copper
grids and imaged using TEM. In general, features of
these ultrathin films were almost identical to those of
their parent homopolymer containing hybrid nanopar-
ticles. In all cases, silica cores (Deff ∼ 20 nm) were well
dispersed in a continuous phase of the tethered copoly-
mers. As shown in SEC of cleaved block copolymers from
hybrid nanoparticles (Table 1 and Figures 7-9), the
dimensions of the outer block segments incorporated to
SiO2-g-pSty, SiO2-g-pBA, and SiO2-g-pMMA hybrid
nanoparticles were relatively short. Thus, visualization
of subtle differences in interparticle spacing was dif-
ficult. Furthermore, the observation of distinct copoly-
mer domains using TEM was not possible without
selective staining of one of the tethered copolymer
segments. Thus, tapping-mode AFM41,54,55 was chosen
to image the effects of grafting block copolymers to silica
cores on the morphology of ultrathin films, as the
presence of both rubbery and rigid domains was antici-
pated to impart sufficient compliance-based contrast.
AFM images of SiO2-g-(pSty175-b-pBA42) (8, Table 1)
ultrathin films revealed a morphology of rigid particles
dispersed in a rubbery matrix. We anticipated that each
1
1.46, Figure 9). H NMR analysis of the product after
chain extension confirmed the incorporation of pMMA
(40 mol %) by extension from the inner pBA segment
(60 mol %). The low molar mass mass tail (Figure 9) in
SEC and the higher polydispersity obtained after chain
extension from SiO2-g-pBA was attributed to the pres-
ence of dead chains formed in the initial ATRP of BA
from the bare silica nanoparticle. The detection of dead
chains in SEC was consistent with kinetic observations
in the ATRP of BA from silica colloidal initiators where
slower rates and formation of excess Cu(II) pointed to
intramolecular termination of surface-immobilized radi-
cals. The average DPn of the cleaved copolymers was
1
determined using H NMR(SiO2-g-(pBA94-b-pMMA62)).
DLS of hybrid nanoparticles in dilute tetrahydrofuran
(THF) solutions confirmed an increase in effective
diameter for SiO2-g-pBA94 (Deff ) 42 nm) and SiO2-g-
(pBA94-b-pMMA62)(Deff ) 60 nm).
Mor p h ology of Ultr a th in F ilm s. The morphology
of 2-bromoisobutyrate-functional silica colloids and
hybrid nanoparticles possessing tethered homopolymers
of pSty, pBA, and pMMA was investigated using TEM
(Figure 10). Samples were prepared by drop-casting
dilute solutions of SiO2 and hybrid nanoparticles onto
carbon-coated copper grids, followed by evaporation of
the solvent in air. Images of SiO2 colloidal initiators
(Figure 10a) indicated that nanoparticles assembled into
two-dimensional aggregates of discrete particles, indi-

Macromolecules, Vol. 36, No. 14, 2003
Organic/Inorganic Hybrid Nanoparticles 5103
F igu r e 11. Tapping mode AFM of an ultrathin film from SiO2-
g-(pS175-b-pBA42) hybrid nanoparticles. SEC of cleaved pSty-
b-pBA from Figure 7, conditions for synthesis, 8, Table 1. Large
spherical features correspond to SiO2-g-pS domains, embedded
in the matrix formed by the outer segments of tethered pBA.
component of the hybrid nanoparticle would be visual-
ized, as for previously prepared SiO1.5-g-(pSty-b-pBzA)
materials.41 From the composition of the tethered
pSty175-b-pBA42 segments and blocking sequence, dis-
persion of rigid domains in a thin rubbery matrix was
expected due to low content of pBA in the material.
Hybrid nanoparticles were spin-coated onto freshly
cleaved mica depositing submonolayer aggregates of
colloids onto the flat surface. Ultrathin films of these
hybrid nanoparticles were deformable due to the pres-
ence of a rubbery outer shell of pBA segments, as rigid
domains were clearly discernible with increasing the
effective tapping force. AFM images of submonolayers
composed of SiO2-g-(pSty175-b-pBA42) colloids (Figure 11)
revealed the presence of spherical particles slightly
spaced in a continuous matrix. Spherical features
visualized in AFM images of these submonolayers were
significantly larger than silica cores imaged in TEM of
SiO2-g-(pSty175-b-pBA42) ultrathin films, indicating that
they corresponded to silica cores surrounded by the
glassy pSty shell. The variation of size of spherical
features seen in Figure 11 presumably reflects the
polydispersity of silica colloidal initiators as previously
shown by TEM (see Figure 10a).
F igu r e 12. Tapping mode AFM images of an ultrathin film
from SiO2-g-(pMMA175-b-pBA79) hybrid nanoparticles acquired
under light tapping (a) and hard tapping (b) conditions. SEC
of cleaved pMMA-b-pBA from Figure 8, conditions for synthe-
sis, 9, Table 1. Large spherical features correspond to SiO2-
g-pMMA domains, while matrix is outer segment of tethered
pBA. Under hard tapping conditions, the soft pBA matrix
became compressed, revealing subsurface SiO2-g-pMMA do-
mains.
AFM images of ultrathin films SiO2-g-(pMMA175-b-
pBA79) (9, Table 1) prepared under similar conditions
as SiO2-g-(pSty175-b-pBA42) also revealed a morphology
consistent with rigid particles dispersed in a rubbery
matrix (Figure 12). The rubbery character of the matrix
was apparent when the tip-sample force was increased
from “light tapping” (Figure 12a) to hard tapping
conditions (Figure 12b). As with the previous sample,
the spherical features in AFM images of SiO2-g-
(pMMA175-b-pBA79) were larger than silica cores ob-
served in TEM, indicating that they corresponded to
silica cores surrounded by glassy pMMA shells. How-
ever, the interparticle spacings were considerably larger
than for the SiO2-g-(pSty175-b-pBA42) system, presum-
ably due to the higher DP of pBA in the outer shell.
F igu r e 13. Tapping mode AFM images of an ultrathin film
from SiO2-g-(pBA94-b-pMMA352) hybrid nanoparticles acquired
under light tapping (a) and hard tapping (b) conditions. Under
hard tapping conditions, the soft pBA shells sandwiched
between rigid cores and rigid pMMA matrix appeared as
characteristic depressed halos.
Of particular interest in AFM analysis was the SiO2-
g-(pBA94-b-pMMA62) system (10, Table 1) with the soft
phase (pBA) sandwiched between the silica core and
glassy shell (pMMA). However, the AFM images of
ultrathin films of these particles prepared by spin-
coating on mica did not reveal the expected presence of
more compliant zones separating the silica cores from
the glassy matrix. This was attributed to the low DP of
the outer pMMA segment, preventing it from segregat-
ing into a continuous rigid matrix upon casting onto
mica. To test this hypothesis, SiO2-g-(pBA94-b-pMMA62)
hybrid nanoparticles were further chain extended with
MMA (11, Table 1) to increase the DP of the pMMA
outer shell. SEC of the cleaved pBA-b-pMMA copolymer
indicated that while low molar mass impurities in-
creased the overall polydispersity of the pBA-b-pMMA
block copolymer (Mn SEC pBA-b-pMMA ) 14 200; Mw/Mn )
3.10), the major product formed a high molar mass peak

5104 Pyun et al.
Macromolecules, Vol. 36, No. 14, 2003
(21) von Werne, T.; Patten, T. E. J . Am. Chem. Soc. 1999, 121,
7409.
(22) von Werne, T.; Patten, T. E. J . Am. Chem. Soc. 2001, 123,
from incorporation of pMMA (Mn main peak pBA-b-pMMA
)
39 100; Mw/Mn ) 1.29). 1H NMR analysis also confirmed
an increase from 40 to 80 mol % in the molar composi-
tion of pMMA after the chain extension reaction. The
final DP of each segment determined from 1H NMR
analysis was SiO2-g-(pBA94-b-pMMA352).
7497.
(23) Farmer, S. C.; Patten, T. E. Chem. Mater. 2001, 13, 3920.
(24) Carrot, G.; Diamanti, S.; Manuszak, M.; Charleux, B.; Vairon,
J . P. J . Polym. Sci., Part A: Polym. Chem. 2001, 39, 4294.
(25) Savin, D. A.; Pyun, J .; Patterson, G. D.; Kowalewski, T.;
Matyjaszewski, K. J . Polym. Sci., Part B: Polym. Phys. 2002,
40, 2667.
Following the extension of pMMA segment, the AFM
images of ultrathin films of SiO2-g-(pBA94-b-pMMA352
)
(11, Table 1) acquired under light tapping (Figure 13a)
and hard tapping conditions (Figure 13b) revealed the
expected presence of more compliant (and thus de-
pressed in height images) zones of pBA, separating the
rigid cores from the rigid, glassy pMMA matrix.
(26) Mori, H.; Seng, D. C.; Zhang, M.; Mueller, A. H. E. Langmuir
2002, 18, 3682.
(27) Chen, X.; Randall, D. P.; Perruchot, C.; Watts, J . F.; Patten,
T. E.; von Werne, T.; Armes, S. P. J . Colloid Interface Sci.
2003, 257, 56.
(28) Nuss, S.; Bottcher, H.; Wurm, H.; Hallensleben, M. L. Angew.
Chem., Int. Ed. 2001, 40, 4016.
(29) Mandal, T. K.; Fleming, M. S.; Walt, D. R. Nano Lett. 2002,
2, 3.
Con clu sion
Hybrid nanoparticles consisting of silica cores teth-
ered with homopolymers and copolymers were prepared
by the ATRP of styrene and (meth)acrylate monomers
from silica colloidal initiators. Comparative kinetics
studies of Sty, BA, and MMA polymerizations from
surfaces in conjunction with control experiments con-
ducted with the corresponding small molecule initiator,
ethyl 2-bromoisobutyrate, revealed the complex nature
of surface-initiated polymerization. However, despite the
complexity of kinetics, ATRP is suitable for the prepara-
tion of hybrid nanoparticles with tethered (co)polymers
of well-defined composition and degree of polymeriza-
tion. TEM and AFM studies showed the impact of graft
composition on the nanoscale morphology of ultrathin
films of these hybrid nanoparticles.
(30) Ohno, K.; Koh, K.-m.; Tsujii, Y.; Fukuda, T. Macromolecules
2002, 35, 8989.
(31) Zhu, M.-Q.; Wei, L.-H.; Zhang, W.-C.; Du, F.-S.; Li, Z.-C.; Li,
F.-M. Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.)
2002, 43, 68.
(32) Tanke, R.; Kauzlarich, S. M.; Patten, T. E.; Pettigrew, K. A.;
Murphy, D. L.; Thompson, M. E.; Lee, H. W. H. Chem. Mater.,
in press.
(33) Wang, J .-Y.; Chen, W.; Liu, A.-H.; Lu, G.; Zhang, G.; Zhang,
J .-H.; Yang, B. J . Am. Chem. Soc. 2002, 124, 13358.
(34) Tsubokawa, N.; Yoshikawa, S. Rec. Res. Dev. Polym. Sci.
1998, 2, 211.
(35) Vestal, C. R.; Zhang, Z. J . J . Am. Chem. Soc. 2002, 124,
14312.
(36) Kickelbick, G.; Holzinger, D.; Brick, C.; Trimmel, G.; Moons,
E. Chem. Mater. 2002, 14, 4382.
(37) Gu, B.; Sen, A. Macromolecules 2002, 35, 8913.
(38) Costa, R. O. R.; Vasconcelos, W. L.; Tamaki, R.; Laine, R. M.
Macromolecules 2001, 34, 5398.
(39) Zhao, B.; Brittian, W. J . Prog. Polym. Sci. 2000, 25, 677.
(40) Qiu, J .; Charleux, B.; Matyjaszewski, K. Prog. Polym. Sci.
2001, 26, 2083.
(41) Pyun, J .; Matyjaszewski, K.; Kowalewski, T.; Savin, D.;
Patterson, G.; Kickelbick, G.; Huesing, N. J . Am. Chem. Soc.
2001, 123, 9445.
Ack n ow led gm en t. The NSF (for K.M., DMR 00-
90409; for T.K., DMR-9871874; for G.D.P., DMR-
9988451) and the Controlled Radical Polymerization
Consortium and Legacy Fellowship at CMU (for J .P.)
are gratefully acknowledged for financial support.
(42) Miller, P. J .; Matyjaszewski, K. Macromolecules 1999, 32,
8760.
Refer en ces a n d Notes
(43) Matyjaszewski, K.; Patten, T. E.; Xia, J . J . Am. Chem. Soc.
1997, 119, 674.
(44) Weast, R. C., Ed.; Handbook of Chemistry and Physics, 53rd
ed.; The Chemical Rubber Co.: Cleveland, 1972.
(45) Matyjaszewski, K.; Pyun, J .; Gaynor, S. G. Macromol. Rapid
Commun. 1998, 19, 665.
(1) Bates, F. S.; Fredrickson, G. H. Annu. Rev. Phys. Chem. 1990,
41, 32.
(2) Lee, M.; Cho, B. K.; Zin, W. C. Chem. Rev. 2001, 101, 3869.
(3) Matjievic, E. Chem. Mater. 1993, 5, 412.
(4) Xia, Y.; Gates, B.; Yin, Y.; Lu, Y. Adv. Mater. 2000, 12, 693.
(5) Caruso, F. Adv. Mater. 2001, 13, 11.
(6) Ulrich, R.; Du Chesne, A.; Templin, M.; Wiesner, U. Adv.
Mater. 1999, 11, 141.
(7) MacLachlan, M. J .; Manners, I.; Ozin, G. A. Adv. Mater. 2000,
12, 675.
(8) Novak, B. M. Adv. Mater. 1993, 5, 422.
(9) Wen, J .; Wilkes, G. L. Chem. Mater. 1996, 8, 1667.
(10) Webster, O. W. Science 1991, 251, 887.
(11) Wang, J .-S.; Matyjaszewski, K. J . Am. Chem. Soc. 1995, 117,
5614.
(12) Matyjaszewski, K.; Xia, J . Chem. Rev. 2001, 101, 2921.
(13) Kamigaito, M.; Ando, T.; Sawamoto, M. Chem. Rev. 2001, 101,
3689.
(14) Wang, J .-S.; Matyjaszewski, K. Macromolecules 1995, 28,
7901.
(15) Patten, T. E.; Matyjaszewski, K. Adv. Mater. 1998, 10, 901.
(16) Patten, T. E.; Xia, J .; Abernathy, T.; Matyjaszewski, K.
Science 1996, 272, 866.
(17) Patten, T. E.; Matyjaszewski, K. Acc. Chem. Res. 1999, 32,
895.
(46) Angot, S.; Murthy, K. S.; Taton, D.; Gnanou, Y. Macromol-
ecules 2000, 33, 7261.
(47) Wang, J .-L.; Grimaud, T.; Matyjaszewski, K. Macromolecules
1997, 30, 6507.
(48) Davis, K. A.; Paik, H.-j.; Matyjaszewski, K. Macromolecules
1999, 32, 1767.
(49) Kajiwara, A.; Matyjaszewski, K. Macromol. Rapid Commun.
1998, 19, 319.
(50) Kajiwara, A.; Matyjaszewski, K. Polym. J . 1999, 31, 70.
(51) Fischer, H. J . Polym. Sci., Part A: Polym. Chem. 1999, 37,
1885.
(52) Heise, A.; Diamanti, S.; Hedrick, J . L.; Frank, C. W.; Miller,
R. D. Macromolecules 2001, 34, 3798.
(53) Shipp, D. A.; Wang, J .-L.; Matyjaszewski, K. Macromolecules
1998, 31, 8005.
(54) Leclere, P.; Moineau, G.; Minet, M.; Dubois, P.; J erome, R.;
Bredas, J . L.; Lazzaroni, R. Langmuir 1999, 15, 3915.
(55) Yu, M. F.; Kowalewski, T.; Ruoff, R. S. Phys. Rev. Lett. 2000,
85, 1456.
(56) Fischer, H. In Free Radicals in Biology and Environment;
Minisci, F., Ed.; Kluwer Acad. Publ.: Dordrecht, 1997.
(57) Beuermann, S.; Buback, M. Prog. Polym. Sci. 2002, 27, 191.
(18) Matyjaszewski, K. Chem.sEur. J . 1999, 5, 3095.
(19) Coessens, V.; Pintauer, T.; Matyjaszewski, K. Prog. Polym.
Sci. 2001, 26, 337.
(20) Pyun, J .; Matyjaszewski, K. Chem. Mater. 2001, 13, 3436.
MA034188T