Synthesis and multidimensional NMR characterization of PDMS-b-PS prepared by combined anionic ring-opening and nitroxide-mediated radical polymerization

Article abstract of DOI:10.1021/ma0114178

The synthesis of PDMS-b-PS copolymers by anionic and nitroxide-mediated radical polymerization (NMRP) is discussed. Hydride end-capped PDMS is hydrosilylated with an allylic alkoxyamine to form a TEMPO functional macroinitiator. The macroinitiator is reac

Full text of DOI:10.1021/ma0114178

4238
Macromolecules 2002, 35, 4238-4246
Synthesis and Multidimensional NMR Characterization of PDMS-b-PS
Prepared by Combined Anionic Ring-Opening and Nitroxide-Mediated
Radical Polymerization
An d r e’ M. Mor ga n ,† Steven K. P olla ck ,*,† a n d Kebed e Besh a h ‡
Department of Chemistry and the Polymer Science and Engineering Program, Howard University,
Washington, D.C. 20059, and Rohm and Haas Research Laboratories, 727 Norristown Road,
Spring House, Pennsylvania 19477
Received August 7, 2001; Revised Manuscript Received March 13, 2002
ABSTRACT: The synthesis of PDMS-b-PS copolymers by anionic and nitroxide-mediated radical
polymerization (NMRP) is discussed. Hydride end-capped PDMS is hydrosilylated with an allylic
alkoxyamine to form a TEMPO functional macroinitiator. The macroinitiator is reacted with styrene to
form diblock copolymer. NMRP of styrene initiated by the PDMS macroinitiator (Mw/Mn ) 1.22) gave
diblocks with Mw/Mn between 1.28 and 1.35. Microstructure characterization by one (1D) and two (2D)
dimensional solution NMR verified the composition of the copolymer end groups and block junction. LC-
NMR analysis shows that nearly homopolymer free diblocks were obtained by this method. Solid-state
NMR analysis verified a phase-separated morphology.
In tr od u ction
using specially designed apparatuses and high-vacuum
techniques in order to obtain well-defined copoly-
mers.11-13 Also, highly purified monomers and solvents
must be used, which are expensive for large-scale
production. Because of these disadvantages, producing
PDMS-b-PS, and other copolymers synthesized by an-
ionic polymerization, by alternative methods is of cur-
rent interest.14
Recently, there has been a burst of activity in an
alternative method of living polymerization called con-
trolled (“living”) radical polymerization. Several types
of controlled radical polymerization techniques have
been introduced: nitroxide-mediated radical polymer-
ization (NMRP), atom transfer radical polymerization
(ATRP), and reversible addition-fragmentation chain
transfer polymerization (RAFT).14,15 NMRP is of interest
in this paper. NMRP involves using a nitroxide, such
as 2,2,6,6-teteramethylpiperidinyl-1-oxy (TEMPO), to
mediate the polymerization of vinyl monomers. Moad
and Rizzardo introduce the use of TEMPO for controlled
radical polymerization.16 The goal of synthesizing nar-
rowly dispersed polymer using the nitroxide remained
elusive until a procedure reported by Georges et al.17
This group synthesized polystyrene using a bimolecular
mixture of benzoyl peroxide and TEMPO, resulting in
polymer with a polydispersity < 1.3. Craig Hawker
developed this method even further by synthesizing a
unimer, 1, by a similar technique. He then used this
Poly(dimethylsiloxane)-b-polystyrene copolymers have
been of scientific and technological interest for many
years.1 Composed of monomers with extremely different
characteristics, these materials exhibit interesting prop-
erties, making them desirable engineering and com-
modity materials. The unique properties of PDMS-b-
PS copolymers are mainly due to the presence of the
inorganic, polysiloxane sequence. Some of the notable
properties of siloxane polymers are high flexibility,
causing them to have very low glass transition temper-
atures (Tg), low surface tension and surface energy, high
gas permeability, and chemical and physiological inert-
ness. Organic-siloxane copolymers such as PDMS-b-
PS have potential uses as thermoplastic elastomers,
compatibiizers, or surfactants in polymer blends. PDMS/
PS block and graft copolymers have been incorporated
into contact lenses, gas permeable films and tubes,
release surfaces for pressure-sensitive adhesives, and
lithographic printing masters.2 PDMS/PS block copoly-
mers have also been used as steric stabilizers during
dispersion polymerization of styrene in supercritical
CO2.3
Poly(dimethylsiloxane)-b-polystyrene copolymers are
traditionally synthesized by “living” anionic polymeriza-
tion.4-6,8-10 More than 30 years ago, Saam et al.4
described the sequential anionic polymerization of
styrene and hexamethylcyclotrisiloxane (D3), using an
alkyllithium initiator. Since then, others have gone on
to synthesize multiple block and graft copolymers of
PDMS and PS via anionic polymerization.6,7,10,11 While
anionic polymerization is successful, it does have some
limitations. In general, anionic polymerization is limited
by its applicability to a small number of monomers and
incompatibility with desired polymer end groups. Also,
reaction conditions call for the strict exclusion of
impurities, such as moisture, oxygen, and CO2. Because
of these requirements, laboratory synthesis is performed
unimer to synthesis controlled molecular weight, nar-
rowly dispersed polymer. Compound 1, containing only
one reactive moiety, was later modified to produce the
† Howard University.
‡ Rohm and Haas Research Laboratories.
10.1021/ma0114178 CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/24/2002

Macromolecules, Vol. 35, No. 11, 2002
Poly(dimethylsiloxane)-b-polystyrene 4239
Sch em e 1
difunctional initiator, 2.18 Kobatake et al.19 used 2 to
synthesize a nitroxy-functionalized polybutadiene mac-
roinitiator. This macroinitiator was then used for po-
lymerization of styrene via NMRP, resulting in polyb-
utadiene-b-polystyrene copolymer.20 A variety of multi-
functional initiators such as 2 have since been used to
create block and graft copolymers by a hybrid of NMRP
with other polymerization techniques (anionic, cationic,
condensation, etc.).14,21-27 Although one step in the
synthesis may still involve ionic polymerization, im-
proved control over molecular design may be gained
using a two-step process.
ionic ring-opening polymerization and nitroxide-medi-
ated radical polymerization (Scheme 1). Thorough char-
acterization of the polymers’ microstructure is performed
using 1D (1H, 13C, and DEPT-135) and 2D (hetero-
nuclear single quantum coherence (HSQC), hetero-
nuclear single quantum coherence-total correlation
spectroscopy (HSQC-TOCSY), and heteronuclear mul-
tiple bond correlation (HMBC) NMR. On-line LC-NMR
has also been employed in order to determine the
presence of homopolymers in diblock samples. Finally,
preliminary morphology studies have been performed
using solid-state NMR.
Two groups have previously described combined “liv-
ing” anionic and NMRP of poly(dimethylsiloxane)-b-
polystyrene. Yoshida and Tanimoto used a PDMS
macroinitiator, containing an azo moiety, with MTEM-
PO (4-methoxy-2,2,6,6-tetramethylpiperidine-1-oxy) to
form PDMS-b-PS, having MTEMPO as one of the end
groups.26 Miura and co-workers used a difunctional
alkoxyamine, 4-lithiophenylalkoxyamine, to initiate
anionic ring-opening polymerization of hexamethylcy-
clotrisiloxane, followed by NMRP of styrene to form the
diblock.28 While not in the scope of this text, PDMS/PS
block and graft copolymers have also been synthesized
by ATRP, using benzyl chloride terminated PDMS to
initiate styrene polymerization.29
Exp er im en ta l Section
Ma ter ia ls. Commercially available hexamethylcyclotrisi-
loxane (Aldrich) was purified by melting and stirring over
CaH2 for ca. 12 h, followed by vacuum sublimation. sec-
Butyllithium (1.3 M in hexane, Aldrich) was titrated using the
Gilman double-titration method.29 Allyl bromide, NaH, (CH3)2-
ClSiH (Aldirch), and the catalyst Pt-divinyltetramethyldisi-
loxane complex (3-3.5% Pt, Gelest) were all used as received.
THF, DMF, styrene, and toluene were purified according to
procedures described in the literature.30
In str u m en ta tion . 1D and 2D NMR characterizations were
performed on two spectrometers. 1H, HSQC-gs, HSQC-TOCSY-
gs, HMBC-gs, and COSY-gs were performed on a Bruker
DMX400 spectrometer, using a 5 mm inverse probe. 13C and
DEPT-135 spectra were obtained using a Bruker AMX500
spectrometer.
We have now added to the list of methods used to
synthesize poly(dimethylsiloxane)-b-poly(styrene) co-
polymers (see Scheme 1). Starting with a difunctional
allyl ether derivative of the alkoxyamine 2, we have
synthesized diblock copolymers of PDMS-b-PS via an-
LC separation of the polymer was performed using a
conventional LC system, equipped with both a UV (wavelength

4240 Morgan et al.
Macromolecules, Vol. 35, No. 11, 2002
220 nm) detector and an evaporative light scattering detector
(ELSD, Polymer Laboratories, Inc.). Reversed phase separation
was done using a SUPLECOSIL C18 column (25 cm × 4.6
mm), with a solvent gradient of acetonitrile/THF from 80/20
to 20/80 in 40 min. The LC system is online with a Varian
UNITY-INOVA 600 MHz NMR spectrometer, with a 60 µL
triple-resonance flow probe, with stop-flow capabilities. A
needle valve splitter immediately after the column splits the
mobile phase between the two detectors, with 85-95% of the
eluent going to the UV detector and subsequently the NMR
1
flow probe; the remaining eluent goes to the ELSD. H NMR
analysis of the copolymer was performed, using pulse field
gradient techniques to suppress signals from solvent and
reaction byproducts.
Syn th esis of BST a n d Ha w k er ’s In itia tor . Compound 1
was synthesized according to the procedure described by
Kobatake, Harwood, and Quirk.19 For purification 1 was
filtered over a column of silica gel and recrystillized from
ethanol to obtain the product in 50% yield based on TEMPO
(mp 72-73 °C). Compound 2 was synthesized according to the
literature.18,19 To ensure complete drying, 2 was dissolved in
benzene, and rotary evaporated to remove the solvent. This
procedure was repeated three times. The product was placed
on a vacuum line to remove residual solvents. 2 was a pale
yellow liquid, obtained in 87% yield.
Syn th esis of Allyl Eth er . Compound 2 was mixed with 4
equiv of NaH in DMF (total concentration 0.1 M of 2 in DMF).
The solution was allowed to stir under N2 for 1 h at room
temperature. After 1 h, 6 equiv of allyl bromide was added
dropwise via an addition funnel, and the solution was allowed
to stir for 20 h at room temperature. The reaction mixture was
diluted with H2O, extracted with hexane, and dried over
MgSO4. A yellow, viscous liquid was obtained, which crystal-
lized after purification by column chromatography (mp 32-
33 °C). 3 was obtained in 73 and 78% product yields.
1
F igu r e 1. (a) H and (b) 13C NMR of allyl ether.
Resu lts a n d Discu ssion
Syn th esis of Hyd r id e En d -Ca p p ed P DMS. Hexameth-
ylcyclotrisiloxane (D3) was dissolved in THF (ca. 2.3 M), and
the solution was stirred under N2 until D3 completely dis-
solved. sec-Butyllithium was then added via syringe to initiate
polymerization. The reaction mixture was allowed to stir under
N2 at ca. 50 °C. To minimize backbiting and chain scrambling
reactions, which broaden molecular weight distribution, the
reaction was monitored by MALDI-ToF MS.31-33 Polymeriza-
tion was quenched with excess (CH3)2ClSiH after 3 h. The
reaction mixture was poured into an excess of acetonitrile to
precipitate the polymer. The polymer (4) was recovered by
decantation and placed on a vacuum line to remove residual
solvents. A colorless, viscous oil was obtained on the average
of ca. 80% yield based on D3.
Hyd r osilyla tion of Hyd r id e En d -Ca p p ed P DMS. Hy-
dride end-capped PDMS (4) was dissolved in toluene (ca. 0.01
M solution) in a pressure bottle. To this, 1.25 equiv (based on
the estimated moles of Si-H in 4) of 3 were added, and the
solution was stirred with a glass rod until 3 completely
dissolved. Next, 2 drops (ca. 10 µL) of platinum catalyst (Pt-
divinyltetramethyl disiloxane complex) were added to the
pressure bottle. The bottle was sealed tightly and placed in
an oil bath at ca. 50-60 °C for 24 h. Upon cooling, the solution
was diluted with toluene and filtered over a plug of deactivated
charcoal and Celite to remove any particulates and for decol-
orization. The polymer was then precipitated from toluene with
methanol and decanted, and the recovered polymer was placed
on a vacuum line to remove residual solvents. A viscous,
slightly murky oil was obtained in ca. 55-60% yield.
Allyl Eth er . The functionality of the alkoxyamine 2
can be altered by reacting it with an allyl halide. The
product is a difunctional alkoxyamine, capable of hy-
drosilylation and nitroxide-mediated radical polymeri-
zation.
1
The H and 13C spectra of 3 are illustrated in Figure
1. Identification of the proton and carbon moieties of
1
this molecule was made using H, 13C, DEPT-135, and
2D HSQC-NMR. The broad multiplet, resonating at
0.5-1.7 ppm, is due to TEMPO. The line width of this
signal is inversely proportional to the value of T2, the
spin-spin relaxation time, which is effected by motional
averaging in a molecule. Lack of motional averaging
causes T2 to be short, resulting in dipolar broadening.
Molecules with restricted motion (i.e., polymers, rigid
solids) have short T2 values, causing broad signals in
the 1H spectrum of the molecule. Hence, the broad
signal in the spectrum of the allyl ether indicates that
TEMPO is locked in a rigid conformation.
In the 13C spectrum of 3, note that the methyl
carbons, 8 and 9, have different chemical shifts. This,
again, alludes to the locked conformation of TEMPO.
As seen with other six-membered cyclic systems (i.e.,
cyclohexane at low temperatures), the axial position
experiences a slightly more shielded magnetic environ-
ment from the equatorial position, causing a difference
in chemical shift. Molecular modeling of the allyl ether
shows that TEMPO in its lowest energy state exists in
a locked chair conformation. Also, carbons 8 and 9’s
proximity to the chiral benzylic carbon 5 attributes to
the chemical shift difference. These factors cause a
chemical shift difference of 13.66 ppm for carbons 8 and
9.
Syn th esis of P DMS-b-P S. A high-pressure reaction tube
was charged with monomer and macroinitiator. Three freeze-
dry cycles were performed to evacuate the reaction tube.
Reaction tubes were then placed in a sand bath for 72 h at
130 °C. After 72 h, the viscous polymer slurry was dissolved
in THF in order to remove it from the reaction tube. An aliquot
of the polymer solution was taken to determine percent
conversion by 1H NMR. The polymer was precipitated from
THF using a large excess of methanol, collected by decantation,
and dried in vacuo. The average conversion was 94%.
Hyd r id e En d -Ca p p ed P DMS. Anionic polymeriza-
tion of hexamethylcyclotrisiloxane (D3) has been exten-

Macromolecules, Vol. 35, No. 11, 2002
Poly(dimethylsiloxane)-b-polystyrene 4241
F igu r e 3. 1H NMR of macroinitiator.
F igu r e 2. 1H NMR of PDMS.
sively reviewed in the literature.32-35 One of the well-
known features of siloxane polymerization is the
formation of cyclic byproducts, due to backbiting and
chain redistribution reactions. The amount of cyclics
formed increases with conversion, until equilibrium is
reached.32-35 This results in a broadened molecular
weight distribution. To prevent this from occurring, the
reaction was monitored by MALDI-ToF mass spectrom-
etry.31 MALDI-ToF analysis of polymer samples taken
from the reaction at different time intervals (30 min, 3
h, 6 h, 9 h, and 24 h) showed that these side reactions
can be kept to a minimum if the reaction is quenched
at 3 h. Allowing polymerization to proceed for 3 h at ca.
50 °C resulted in 80% yield based on the amount of
monomer used.
Ta ble 1. Resu lts of Hyd r id e En d -Ca p p ed P DMS a n d
Ma cr oin itia tor
Mn
Mn
Mw
Mw/Mn
sample ID
PDMS (4)
macroinitiator (5)
[D3/s-BuLi] (calc) (GPC) (GPC) (GPC)
16
na
3749 4606
4066 4167
5552
5094
1.20
1.22
Ta ble 2. Resu lts of Diblock Syn th esis
[Sty/4]
Mn
Mn
Mw
Mw/Mn
sample ID (molar ratio)
(calc)
(GPC)
(GPC)
(GPC)
DB-1
DB-2
DB-3
154
74
50
20 330 19 754 26 645
11 885 10 378 13 156
1.35
1.27
1.29
9 369
9 495 12 242
P DMS-b-P S. Diblocks were formed by bulk stable
free radical polymerization of styrene, using the mac-
roinitiator as the initiator and nitroxide source. Table
2 summarizes the GPC results of the diblocks. For each
diblock, approximately equal amounts of macroinitiator,
4, were used, varying the amount of styrene monomer.
Comparing the Mn values for diblocks 1, 2, and 3, Mn
increased with an increasing amount of monomer
charged to the reaction tube. This indicates that the
reaction is behaving in a living manor, with styrene
polymerization initiated by the macroinitiator. For
diblocks 1 and 3, Mn values determined from GPC are
less than 5% different from the calculated Mn’s. Diblock
2 shows a slight deviation from the calculated Mn, but
results are still comparable to the theoretical value.
GPC elution curves for these samples show unimodal
distributions, with the elution curve of diblock 1 having
a slight shoulder at the high molecular weight region.
This indicates some inhomogeneity in diblock 1, result-
ing from autopolymerization of styrene, or diblocks very
heavy in styrene content (see LC-NMR Characteriza-
tion). For all three samples, Mw/Mn e 1.35, which is
below the theoretical limiting polydispersity of 1.5 for
a conventional free-radical polymerization process. With
only a slight increase in the polydispersity of the
diblocks as compare to the macroinitiator, one may
assume that the PS block is nearly monodispersed.
However, it is realized that GPC (based on PS stan-
dards) cannot be used to determine the polydispersity
for each individual block. Other techniques such as
LCCC or MALDI-ToF MS must be used to evaluate
these properties.
1
The H NMR spectrum of 4 is illustrated in Figure 2.
An intense peak resonating at 0.074 ppm represents the
dimethyl protons on Si. The end groups, sec-butyl and
Si-H, resonate at 0.51-1.59 and 4.72 ppm, respectively.
The presence of the Si-H functional group was also
confirmed by FT-IR, 2128.0 cm-1. Comparing the inte-
grals of the sec-butyl peaks and the Si-H peak results
in a 9:1 ratio. This indicates that the majority of the
polymer chains have the same end groups.
Ma cr oin itia tor . Hydrosilylation of hydride end-
capped PDMS results in the macroinitiator 5. The
polymer was recovered by precipitation, leaving any
unreacted allyl ether and/or nonpolymeric byproducts
in solution. Confirmation of hydrosilylation was first
made by FT-IR. The Si-H absorption band, at 2128.0
cm-1, was absent from the spectrum of the macroini-
tiator upon completion of the reaction.
1
The H NMR spectrum is illustrated in Figure 3. The
presence of TEMPO is indicated by the characteristic
broad signal, resonating at 0.4-1.6 ppm. Signals due
to the sec-butyl end group lie underneath the broad
TEMPO multiplet. Comparing the 1H spectra of the allyl
ether to that of the macroinitiator (see Figures 1a and
3), one should notice the disappearance of the allylic
proton peaks at 5.13 and 5.77 ppm, further indicating
hydrosilylation. The CH2 protons 3 have also shifted
upfield to 3.3 ppm (f in Figure 3). Peaks corresponding
to 4, 5, and 6 (g, h, and i in Figure 3) remain at roughly
the same chemical shift. There is no indication of side
reactions, such as R-addition, or unreacted hydride end-
capped PDMS in the spectrum.

4242 Morgan et al.
Macromolecules, Vol. 35, No. 11, 2002
F igu r e 6. HSQC-TOCSY NMR of PDMS-b-PS.
F igu r e 4. 1H NMR of PDMS-b-PS.
or methine carbons neighboring a phenyl group; how-
ever, their exact identity could not be determined by
HSQC. Cross-peaks f, g, and k correlate to methylene
and methine carbons at 73.87, 78.99, and 85.55 ppm.
On the basis of the chemical shift, these carbons are
alkoxy in nature.
Next, connectivity between neighboring groups is
established by HSQC-TOCSY. HSQC-TOCSY shows
1J CH 1H-13C correlation, along with H-1H correlations
1
(TOCSY). In an HSQC-TOCSY spectrum, one observes
the same cross-peaks from HSQC, with new cross-peaks
arising due to TOCSY. Direct correlations between
neighboring moieties can be determined by this tech-
nique, enabling further elucidation of the polymer
microstructure.36 Figure 6 illustrates the HSQC-
TOCSY spectrum of the diblock. Connectivity is estab-
lished by finding correlations between cross-peaks
identified by HSQC to new cross-peaks due to TOCSY.
For example, the methylene carbon d at 14.03 ppm has
two cross-peaks: one present at 0.398 ppm seen in the
HSQC spectrum and a new cross-peak at 1.45 ppm due
to TOCSY. These cross-peaks correlate this carbon to
its directly attached protons at 0.398 ppm and the
neighboring methylene protons at 1.45 ppm. This same
connectivity can be established by carbon e at 23.29
ppm, which also has cross-peaks at 0.398 and 1.45 ppm.
Hence, carbons d and e are connected. Other connec-
tivities, established in a similar fashion, are as follows:
aCH3-bCH-cCH2-a′CH3, representing the sec-butyl end
F igu r e 5. HSQC NMR of PDMS-b-PS.
NMR Ch a r a cter iza tion . 1H NMR verifies gross
monomer composition. In Figure 4, the sharp peak
resonating at 0.074 ppm represents the PDMS block of
the polymer. Broad peaks at 1.2-2.5 ppm (aliphatic
backbone) and 6.0-7.5 ppm (aromatic side chain) rep-
resent the PS block. Peaks due to the polymer micro-
structure are observed between 0.3 and 5.0 ppm.
Microstructure characterization is carried out using
multidimensional NMR techniques. First, 1J CH correla-
d
group, and CH2-eCH2-fCH2-O, representing part of
1
tions are established by HSQC. A H subspectrum of
the block junction. The methylene alkoxy carbon g
shows signs of correlation with a cross-peak at around
2.4 ppm in the 1H dimension. Connectivity between
these moieties was confirmed by COSY, showing cor-
relation between protons g and h. For TEMPO, the only
correlation that can be detected by HSQC-TOCSY is
between its methylene groups. From 13C analysis of the
allyl ether, methylene carbon 10 of TEMPO resonates
in the same region as the aliphatic PS signal in the
diblock. Hence, connectivity within TEMPO cannot be
confirmed, due to the overlap of the aliphatic PS signal.
To confirm connectivity within TEMPO, HMBC was
performed. HMBC was chosen, because it allows one to
the aliphatic region is illustrated on the x-axis of the
HSQC spectrum (Figure 5). DEPT-135 was used to
determine the nature of each carbon moiety, and this
subspectrum is illustrated on the y-axis. Cross-peaks a
and a′ correspond to methyl carbons at 13.03 and 13.35
ppm. The cross-peaks d and n correspond to methylene
peaks at 14.09 and 17.16 ppm. Methyl carbon peaks at
20.24 and 33.90 ppm, o and p, are assigned to two cross-
peaks each. These cross-peaks indicate carbons with
nonequivalent protons. The methylene peak c is also
assigned to two cross-peaks, again indicating non-
equivalent protons. Overlapping peaks b and e cor-
respond to a methine and a methylene group with
approximately the same chemical shift (23.29 ppm). In
the region where the aliphatic PS signal occurs, there
are three cross-peaks that do not correspond to the
polymers backbone. These peaks are due to methylene
establish 1H-13C connectivity via multiple bond J CH
,
even through nonpotonated heteroatoms. The HMBC
spectrum is illustrated in Figure 7. In this 2D spectrum
the 13C spectrum is substituted for DEPT-135 in order
to illustrate correlations with nonprotonated carbons.

Macromolecules, Vol. 35, No. 11, 2002
Poly(dimethylsiloxane)-b-polystyrene 4243
TEMPO moieties, J CH would have to be detected through
a minimum of four bonds (4J CH), exceeding the detection
limits under the current acquisition parameters for this
experiment. Despite the fact that connectivity between
1
k and TEMPO was not established by HMBC, the H
chemical shift of TEMPO-bonded k proton, resonating
at 3.9 ppm (see Figure 5), agrees with values reported
in the literature.37,38 Proton and carbon chemical shifts
for the identified groups are illustrated in Figure 8 and
summarized in Table 3.
Now that the polymer’s microstructure has been
verified, the relative abundance of each end group must
be discussed. Comparing the integral of carbon peak n
of TEMPO to that of c of the sec-butyl group, the ratio
is approximately 2:3. If all polymer chains contain the
TEMPO end group, the relative ratio should be 2:1. The
low abundance of TEMPO as compared to the sec-butyl
end group indicates that all polymer chains are not
capped with TEMPO. It has been reported that other
end groups may result during NMRP of styrene due to
hydrogen-transfer reactions.39 Hydrogen transfer results
in an unsaturated -CHdCHPh end group from dispro-
portionation and a saturated -CH2-CH2Ph end group
from oxidation of the benzylic radical with hydroxy
TEMPO, formed during propagation. There is no evi-
dence of the unsaturated end group from NMR analysis.
There is, however, some evidence of aliphatic protons
from COSY. Unfortunately, their exact identity and
connectivity were not determined.
F igu r e 7. HMBC NMR of PDMS-b-PS.
In the HMBC spectrum, one observes cross-peaks cor-
responding to carbons l (quaternary carbon of TEMPO,
absent from DEPT-135) and p around 59.0 and 33.9
ppm, respectively. Also, there is a cross-peak at 40.5
ppm in the 13C dimension, corresponding to a carbon
underneath the aliphatic PS signal, m. Cross-peaks l,
m, and p occur at the same proton chemical shift as the
2
methyl protons o (see Figure 5), establishing J CH and
3J CH connectivity between o with l, m, and p. Cross-peak
k shows connectivity to aromatic protons, establishing
its benzylic character. However, no connectivity between
k and TEMPO is seen by HMBC. To observe 1H-X
correlation between the benzylic carbon k and any
LC-NMR Ch a r a cter iza tion . Autopolymerization of
styrene is known to occur during NMRP.39-41 To scout
for homopolymer PS and unreacted PDMS homopoly-
mer in the sample, LC-NMR was performed. Figure 9
F igu r e 8. 1H and 13C subspectra of PDMS-b-PS.

4244 Morgan et al.
Macromolecules, Vol. 35, No. 11, 2002
Ta ble 3. 13C a n d 1H Ch em ica l Sh ifts, Deter m in ed by 1D a n d 2D NMR
sec-butyl end group
block junction
13C in ppm
TEMPO end group
13C in ppm
position
δ
13C in ppm
δ1H in ppm
position
δ
δ 1H in ppm
position
δ
δ 1H in ppm
a
a′
b
13.03
13.35
23.29
0.948
0.948
0.542
d
e
f
14.09
23.29
73.57
0.398
1.451
3.159
j
k
l
n.i.
85.55
58.76, 59.84
n.i.
3.998
NA
c
24.04
1.143, 1.560
g
h
75.80
n.i.
3.260
2.401
m
n
o
40.48
17.16
20.24
n.i.
1.158
1.083
0.913
0.937
0.401
p
33.90
F igu r e 10. 1H NMR results from LC-NMR of DB-1.
was detected for one of the samples, DB-3. In Figure
11, LC analysis using an ELSD detector showed a peak
at 26.8 min. The lack of UV response at this retention
time indicates that this peak is due to homopolymer
PDMS. This is confirmed by 1H NMR, having the
characteristic PDMS peak around 0.1 ppm. Compared
to the ELSD response of the diblock, this peak is very
minor. The minute presence of PDMS homopolymer
may be due to incomplete reaction of the hydride-
terminated PDMS. Despite the detection of trace
amounts of PDMS in DB-3, it can be concluded that
these materials are nearly free of homopolymers.
Although LC-NMR results show that these diblocks
were nearly free of homopolymers, there was some
inhomogeneity in these samples. These differences are
further indicated by the change in the relative intensity
of the PS and PDMS peaks with retention time. For
DB-1, the relative molar ratio of PS to PDMS was
estimated from integration of peaks in Figure 10. Based
on integration, the molar ratio of PS:PDMS varied from
3:1 to 6:1. The maximum ratio was approximately 6:1
at 18.08 min, the highest intensity region of the chro-
matograph. The minimum PS:PDMS molar ratios were
F igu r e 9. LC of diblocks (a) DB-1, (b) DB-2, and (c) DB-3.
illustrates LC (UV detector) data for diblocks 1, 2, and
3 (DB-1, DB-2, and DB-3). In each chromatograph,
there is a broad peak, with some regions of inhomoge-
1
neity. Figure 10 illustrates H NMR results for diblock
1 between 16.0 and 23.0 min. At each retention time,
the corresponding spectrum illustrates a sharp PDMS
peak around 0.1 ppm and a broad PS peak at 6.0-7.5
ppm, proving that block copolymer exist between these
retention times. In both parts b and c of Figure 10, there
exist a shoulder just below 16.0 min that is inconsistent
1
with the main peak. H NMR data for this region are
not shown; however, results for this region indicate that
this shoulder is block copolymer. Also, LC analysis of
polystyrene standards under the same conditions showed
that polystyrene elutes between 7.0 and 14.0 min.
Hence, no homopolymer polystyrene is detected in any
of the samples. These results indicate that any autopo-
lymerization of styrene that may have occurred during
copolymer synthesis in minimal. PDMS homopolymer

Macromolecules, Vol. 35, No. 11, 2002
Poly(dimethylsiloxane)-b-polystyrene 4245
F igu r e 12. Solid-state NMR (cphp TOSS-WISE) spectra of
DB-1.
F igu r e 11. (a) LC (ELSD) for DB-3. (b) 1H spectrum at
retention time of 26.9 min.
in order to determine the specific morphology of the
diblocks.
4:1 and 3:1 at 16.04 and 20.58 min, respectively. 1H
NMR results for DB-2 and DB-3 also indicated a change
in relative intensity. Ratios of PS:PDMS varied from 5:1
to 1:1 in these samples, with the molar ratios being
approximately 2:1 in the highest intensity region of the
chromatograph for both samples. The shoulder just
below 16.0 min has the maximum ratio of PS to PDMS,
corresponding to 5:1 and 4:1 in DB-2 and DB-3,
respectively. These results indicate that regions of
inhomogeneity in the chromatographs are due to diblocks
with varying PS to PDMS molar ratios.
Solid -Sta te NMR. One of the aims of synthesizing
PDMS-b-PS by this two-step process is to use this
technique to make diblocks with specific morphologies.
To obtain a specific morphology (lamellar, cylinders,
spheres, bicontinueous), the system of interest must
exhibit phase segregation between the two blocks.
Considering the drastically different Tg’s of PDMS (Tg
-125 °C) and PS (Tg 104 °C, depending on molecular
weight), one would certainly expect diblocks of these
materials to be phase segregated. To verify this, a WISE
solid-state NMR (cphp-TOSS WISE) experiment was
performed. This is a 2D solid-state NMR experiment,
which correlates carbon and proton chemical shifts.
Using the WISE experiment, one may distinguish
between carbons attached to protons in a rigid environ-
ment vs those in a mobile environment.42 As explained
previously, materials with restricted motion have broad
protons signals, while mobile molecules have relatively
sharp signals. Figure 12 illustrates the WISE spectrum
of DB-1. The sharp signal in the proton dimension (y-
axis) is correlated with the PDMS carbon signal (x-axis),
while the broad line width correlates to the PS carbon
signals. These correlations indicate that there exists a
hard and soft segment of the diblock, verifying phase-
segregated morphology.
Con clu sion
PDMS-b-PS copolymers have been synthesized by
anionic polymerization and nitroxide-mediated radical
polymerization. Molecular weights determined by GPC
agreed well with calculated values, yielding relatively
low polydispersities. Microstructure characterization
verified the incorporation of the sec-butyl and TEMPO
end groups and the block junction. LC-NMR analysis
showed that nearly homopolymer free diblock could be
obtained by this method. Solid-state NMR analysis
verified the phase-segregated morphology, typical of
PDMS/PS copolymers. From the information presented,
one may conclude that this is a viable technique for
synthesizing PDMS-b-PS copolymers. Diblocks of a
specific architecture, with reasonable control over the
molecular weight and polydispersity, can be obtained
without the use of high-vacuum techniques and special
equipment necessary for anionic polymerization. How-
ever, one may question the efficiency of this method
since multiple steps are involved. A more efficient mode
of polymerization is currently being explored. This
method involves synthesizing TEMPO end-capped PDMS
using a lithium alkoxide of the Hawker initiator, fol-
lowed by NMRP of PS. Diblocks formed by this route
will be discussed in a future publication.
Ack n ow led gm en t. Special thanks to Rohm and
Haas Co. for the support of this research through the
Minority Graduate Research Internship Program
(MGRIP). A.M.M. thanks Dr. J ian Wu and Dr. Dan
Dohmeier of the Rohm and Haas Co. for discussions and
applications of LC-NMR and preparative GPC.
Refer en ces a n d Notes
(1) (a) Yilgor, I.; Mcgrath, J . E. In Polysiloxane Copolymers/
Anionic Polymerization; Yilgor, I., McGrath, J . E., Eds.;
Advances in Polymer Science 86; Springer-Verlag: New York,
1988; pp 1-86. (b) Mark, J . In Silicon-Based Polymer
Science: A Comprehensive Resource; Zeigler, J . M., Fearon,
F. W., Gordon, Eds.; Advances in Chemistry Series 224;
TEM analysis of these materials was attempted but
unsuccessful due to difficulty in casting films from
common solvents.43 Higher molecular weight materials,
with the molecular weight of each component above its
entanglement molecular weight, must be synthesized

4246 Morgan et al.
Macromolecules, Vol. 35, No. 11, 2002
American Chemical Society: Washington, DC; pp 47-70. (c)
Chojnowski, J . In Siloxane Polymers; Clarson, S. J ., Semlyen,
J . A., Eds.; Polymer Science and Technology Series; PTR
Prentice Hall: Engelwood Cliffs, NJ , 1993; pp 1-62.
(2) (a) Yokota, M.; Goshima, T.; Fujioka, S. US Patent 4,632,-
968, 1986. (b) Falender, J . R.; Lindsey, S. E.; Ward, A. H.;
Kendrick, T. C. US Patent 3,975,455, 1976. (c) Mann, R. H.;
Sun, E. I.; Plamthottam, S. S.; Newing, C. W. US Patent 5,-
728,469, 1998. (d) Pacansky, T. J .; Amidon, A. B. US Patent
4,218,514, 1980. (e) Gabor, A. H.; Ober, C. K. In Microelec-
tronics Technology. Polymers for Advanced Imaging and
Packaging; ACS Symposium Series 614; American Chemical
Society: Washington, DC, 281.
(3) (a) DeSimone, J . M.; J ones, T. A.; Shaffer, K. A. Macromol-
ecules 1996, 29, 2704. (b) DeSimone, J . M.; Betts, D. E.;
Canelas, D. A. Macromolecules 1996, 29, 2818. (c) DeSimone,
J . M.; Canelas, D. A. Macromolecules 1997, 30, 5673.
(4) Saam, J . C.; Gordon, D. J .; Lindsey, S. Macromolecules 1970,
3, 1.
(5) Zillion, J . G.; Roovers, E. L.; Bywater, S. Macromolecules
1975, 8, 573.
(6) Bajaj, P.; Varshney, S. K.; Misra, A. J . Polym. Sci., Polym.
Chem. Ed. 1980, 18, 295.
(7) Dems, A.; Strobin, C. Makromol. Chem. 1991, 192, 2521.
(8) Chu, J . H.; Rangarajan, P.; Adams, J . L.; Register, R. A.
Polymer 1995, 36, 1569.
(9) Molenberg, A.; Mo¨ller, M.; Pieper, T. Macromol. Chem. Phys.
1998, 199, 299.
(22) Malmstro¨m, E. E.; Hawker, J . Macromol. Chem. Phys. 1998,
199, 923.
(23) Miura, Y.; Hirota, K.; Moto, H.; Yamada, B. Macromolecules
1999, 32, 8356.
(24) Tsoukatos, T.; Pispas, S.; Hadjichristidis, N. Macromolecules
2000, 33, 9504.
(25) Hua, F. J .; Yang, Y. L. Polymer 2001, 42, 1361.
(26) Yoshida, E.; Tanimoto, S. Macromolecules 1997, 30, 4018.
(27) Miura, Y.; Hirota, K.; Moto, H.; Yamada, B. Macromolecules
1999, 32, 8356.
(28) Matyjaszewski, K.; Miller, P. J .; Nakagawa, Y. Polymer 1998,
39, 5163.
(29) Gilman, H.; Carledge, F. K. J . Organomet. Chem. 1964, 2,
447.
(30) Armarego, W. L. F.; Perrin, D. D. Purification of Laboratory
Chemicals, 4th ed.; Butterworth-Heinemann: Oxford, 1996.
(31) Pollack, S. K.; Singer, D. U.; Morgan, A. M. Polym. Prepr.
(Am. Chem. Soc., Div. Polym. Chem.) 2000, 41 (1), 631.
(32) (a) Yilgor, I.; McGrath, J . E. In Polysiloxane Copolymers/
Anionic Polymerization; Yilgor, I., McGrath, J . E., Eds.;
Advances in Polymer Science 86; Springer-Verlag: New York,
1988; pp 1-86. (b) Mark, J . In Silicon-Based Polymer
Science: A Comprehensive Resource; Zeigler, J . M., Fearon,
F. W. G., Eds.; Advances in Chemistry Series 224; American
Chemical Society: Washington, DC; pp 47-70. (c) Chojnows-
ki, J . In Siloxane Polymers; Clarson, S. J ., Semlyen, J . A.,
Eds.; Polymer Science and Technology Series; PTR Prentice
Hall: Engelwood Cliffs, NJ , 1993; pp 1-62.
(33) Boileau, S. In Ring Opening Polymerization; McGrath, J . E.,
Ed.; American Chemical Society: Washington, DC, 1985; pp
23-35.
(34) Lee, C. L.; J ohannson, O. K. Polym. Prepr. (Am. Chem. Soc.,
Div. Polym. Chem.) 1969, 10 (2), 1361.
(35) Veith, C. A.; Cohen, R. E. Makromol. Chem., Macromol.
Symp. 1991, 42, 241.
(36) Beshah, K. Macromolecules 1992, 25, 5597.
(37) Chong, K.; Rizzardo, E.; Solomon, D. H. J . Am. Chem. Soc.
1983, 105, 7761.
(10) Rosati, D.; Perrin, M.; Navard, P. Macromolecules 1993, 31,
4301.
(11) Bellas, V.; Iatrou, H.; Hadjichristidis, N. Marcromolecules
2000, 33, 6993.
(12) Morton, M.; Fetters, L. J . Rubber Chem. Technol. 1975, 48,
359.
(13) Hadjichristidis, N.; Iatrou, H.; Pispas, S.; Pitsikalis, M. J .
Polym. Sci., Part A: Polym. Chem. 2000, 38, 3211.
(14) Matyjaszewski, K.; Teodorescu, M.; Acar, M. H.; et al.
Macromol. Symp. 2000, 157, 183.
(15) Matyjaszewski, K., Ed. Controlled Radical Polymerization;
ACS Symposium Series 685; American Chemical Society:
Washington, DC, 1998.
(16) Moad, G.; Rizzardo, E.; Solomon, D. H. Macromolecules 1982,
15, 909.
(38) Komber, H.; Gruner, M.; Malz, H. Macromol. Rapid Commun.
1998, 19, 83.
(39) He, J .; Li, L.; Yang, Y. Macromolecules 2000, 33, 2286.
(40) Georges, M. K.; Veregin, R. N. P.; Kazmaier, P. M. Trends
Polym. Sci. 1994, 2, 66.
(17) Georges, M. K.; Veregin, R. P. N.; Daimon, K. In Controlled
Radical Polymerization; Matyjaszewski, K., Ed.; ACS Sym-
posium Series 685; American Chemical Society: Washington,
DC, 1998; pp 170-179.
(18) Hawker, C. J .; Hedrick, J . L. Macromolecules 1995, 28, 2993.
(19) Kobatake, S.; Harwood, H. J .; Quirk, R. P. Macromolecules
1997, 30, 4238.
(41) Devonport, W.; Michalak, L.; Malmstro¨m, E.; Mate, M.; Kurdi,
B.; Hawker, C. J .; Barclay, G. G.; Sinta, R. Macromolecules
1997, 30, 1929 and literature cited therein.
(42) Bovey, F. A.; Mirau NMR of Polymers; Academic Press: San
Diego, CA, 1996.
(43) Krause, S.; Iskandar, M.; Iqbal, M. Macromolecules 1982, 15,
105.
(20) Kobatake, S.; Harwood, H. J .; Quirk, R. P. Macromolecules
1998, 31, 3735.
(21) Puts, R. D.; Sogah, D. Y. Macromolecules 1996, 30, 7050.
MA0114178