Effect of mesophase order on the dynamics of side group liquid crystalline polymers

Article abstract of DOI:10.1021/ma050551f

Rheology and X-ray scattering were employed to probe the viscoelastic properties and structural transitions of model cyano-biphenyl-based side-group liquid-crystalline polymers (SGLCPs) with molecular weights ranging from 91 to 1900 kg/mol. Temperature-dependent rheological data show a rapid change in dynamics over a small temperature range. Small-angle X-ray scattering reveals these changes to be associated with an isotropic to smectic transition with an appreciable biphasic region. The presence of a biphasic region is attributed to inhomogeneity in chain structure resulting from incomplete attachment of mesogens to every monomeric unit in the SGLCP polymer. While isotropic and smectic phase data may be separately time-temperature shifted to create master curves for the individual phases, we argue against attempts to achieve superposition between the two phases in the high-frequency regime, since smectic ordering may not simply slow the dynamics but also increase the modulus of the sample. Molecular weight has a strong influence on rheology in the isotropic phase, where an entanglement plateau emerges; however, the smectic-phase rheology is dominated by the layer structure and is fairly insensitive to molecular weight.

Full text of DOI:10.1021/ma050551f

6946
Macromolecules 2005, 38, 6946-6953
Effect of Mesophase Order on the Dynamics of Side Group Liquid
Crystalline Polymers
Maria L. Auad,† Michael D. Kempe,† Julia A. Kornfield,† Stanley Rendon,‡
Wesley R. Burghardt,*,‡ and Kyunghwan Yoon§
Department of Chemistry and Chemical Engineering, California Institute of Technology,
Pasadena, California 91125, Department of Chemical and Biological Engineering,
Northwestern University, Evanston, Illinois 60208, and Department of Mechanical Engineering,
Dankook University, Seoul, Korea
Received March 15, 2005; Revised Manuscript Received June 8, 2005
ABSTRACT: Rheology and X-ray scattering were employed to probe the viscoelastic properties and
structural transitions of model cyano-biphenyl-based side-group liquid-crystalline polymers (SGLCPs)
with molecular weights ranging from 91 to 1900 kg/mol. Temperature-dependent rheological data show
a rapid change in dynamics over a small temperature range. Small-angle X-ray scattering reveals these
changes to be associated with an isotropic to smectic transition with an appreciable biphasic region. The
presence of a biphasic region is attributed to inhomogeneity in chain structure resulting from incomplete
attachment of mesogens to every monomeric unit in the SGLCP polymer. While isotropic and smectic
phase data may be separately time-temperature shifted to create master curves for the individual phases,
we argue against attempts to achieve superposition between the two phases in the high-frequency regime,
since smectic ordering may not simply slow the dynamics but also increase the modulus of the sample.
Molecular weight has a strong influence on rheology in the isotropic phase, where an entanglement plateau
emerges; however, the smectic-phase rheology is dominated by the layer structure and is fairly insensitive
to molecular weight.
1. Introduction
dynamics of entangled SGLCP systems when compared
to the isotropic phase. The rheology of nematic SGLCPs
has continued to attract interest.14-17
Side-group liquid-crystalline polymers (SGLCPs)
consist of stiff mesogenic units laterally attached to a
flexible polymeric backbone via short, flexible spacers.
The partial coupling between the mesogens and the
backbone leads to distinctive macroscopic relationships
between strain and orientation that may find applica-
tion in technologies such as stress and photomechanical
sensors and actuators.1-5 To realize these applications,
it is vital to gain a comprehensive understanding of the
SGLCP rheology as well as the effect that the flow
history has on molecular alignment. It has been found
that the coupled dynamics of the backbone and meso-
gens give rise to intriguing rheological consequences in
the liquid crystalline state.
Early SGLCP melt rheology investigations concluded
that the flexible backbone dominates the dynamics in
nematic SGLCPs, overwhelming typical liquid-crystal-
line characteristics.6-8 These results are surprising,
given that the liquid-crystalline order of small-mole-
cule LCs, lyotropic rodlike LCPs, and thermotropic
main-chain LCPs has a profound effect on their
rheology.9-12
Fewer efforts, however, have examined the viscoelas-
ticity and phase behavior of smectic SGLCPs.6,13,19,20 The
positional order in smectic LCs leads to more pro-
nounced rheological consequences upon order than is
seen in nematic LCs. Colby et al. reported that smectic
SGLCPs exhibit a broad relaxation time distribution
without an accessible terminal regime and document
dramatic changes in dynamics upon transition from the
nematic to smectic phase.6 Rubin et al. observed similar
changes in dynamics upon a nematic to smectic transi-
tion in their study.13 Wewerka et al. employed linear
rheology to probe the viscoelastic properties and struc-
tural transitions in model norbornene-based SGLCPs.19
In their study, SGLCPs with long spacer groups were
found to exhibit smectic ordering at lower temperatures,
in which the characteristic layered morphology domi-
nates the storage modulus.19 More recently, Kim and
Han have compared nematic and smectic SGLCPs and
report much more dramatic changes in linear viscoelas-
ticity for an isotropic-smectic transition than for a
isotropic-nematic transition.20
In this investigation, we report on the synthesis,
phase behavior, and viscoelasticity of a series of smectic
SGLCPs of varying molecular weight. The structural
and rheological properties of these liquid crystalline
materials were investigated using combined rheology
and X-ray techniques to identify and classify the nature
of the mesophase transitions. (A separate paper de-
scribes flow-induced alignment of the smectic layers.21)
We find that the interplay between molecular weight,
polydispersity, and degree of mesogen attachment ex-
plains the origin and width of a smectic-isotropic
biphase.
Investigations of the viscoelasticity of SGLCPs using
steady shear have shown that isotropic-to-nematic
phase transitions have no effect on the melt viscosity.8
Contrary to these results, Rubin et al.13 determined that
nematic order produces significant changes in the
* To whom correspondence should be addressed. E-mail:
w-burghardt@northwestern.edu.
† Department of Chemistry and Chemical Engineering, Cali-
fornia Institute of Technology, Pasadena, California 91125.
‡ Department of Chemical and Biological Engineering, North-
western University, Evanston, Illinois 60208.
§ Department of Mechanical Engineering, Dankook University,
Seoul, Korea.
10.1021/ma050551f CCC: $30.25 © 2005 American Chemical Society
Published on Web 07/16/2005

Macromolecules, Vol. 38, No. 16, 2005
Side-Group Liquid-Crystalline Polymers 6947
Figure 1. Method for synthesis of SGLCP with a siloxane
spacer. For our system m ) 200-2800. The product polymer
is named PBSiCB5 since it was made with a 1,2-polybutadiene
parent polymer, using a siloxane linking group, a cyano-
biphenyl mesogenic unit, and a pentyl group in the spacer.
Figure 2. NMR spectrum of PBSICB5 used to calculate the
fractional attachment of mesogen to the pendant vinyl groups.
The mole fraction of the different monomeric units of mesogen,
1,2-butadiene, and 1,4-butadiene can be calculated from the
peak areas A, B, and C as: fmesogen ) (2A)/(2A + B + 2C), f1,2
) (2B)/(2A + B + 2C), and f1,4 )(2C - B)/(2A + B + 2C),
respectively.
2. Experimental Section
2.1. Polymer Synthesis. The polymers used in this study
were synthesized by mesogen attachment to pendant vinyl
groups of an anionically synthesized22-24 1,2-polybutadiene
backbone (Figure 1). This technique was used so that large
molecular weight polymers could be obtained while still
maintaining relatively low polydispersity. The techniques used
to attach the mesogenic group are similar to those used by
other researchers;25-28 however, due to the large molecular
weight of our polymers, additional effort was required for
reactant and product purification.
The 4′-cyano-4-hydroxylbiphenyl was obtained from TCI
America and purified on a silica gel column using a mixture
of 33% ethyl acetate in hexane. All other chemicals were
purchased from Aldrich and were used without further puri-
fication. The spacer was attached using the Williamson ether
synthesis technique with 30% excess 5-bromopentene in DMF
with 1 equiv of anhydrous potassium carbonate at a temper-
ature of 90 °C for 3 h, Figure 1. The product 4-cyano-4′-(4-
penteneoxy)-biphenyl, CBV5, was precipitated out by the
addition of water and purified on a silica gel column using 50%
dichloromethane in hexane with an overall yield of ∼95%.
A 10× excess of tetramethyldisiloxane (TMDS) dissolved in
anhydrous toluene was used in the next step to reduce the
amount of dimer formed. It was attached to the CBV5 in an
anhydrous toluene solution using a few drops of the catalyst
PC085, platinum-cyclovinyl-methylsiloxane complex, or PC072,
platinum-divinyltetramethyldisiloxane complex in xylene, both
of which were purchased from United Chemical Technology.
This reaction, producing 4-cyano-4′-(5-(1,1,3,3-tetramethyl-
disiloxane)pentoxy) biphenyl (CBSi5), ran overnight at room
temperature and was monitored using thin-layer chromatog-
raphy.
Since only a small amount of disilane byproducts are
necessary to cross link a large molecular weight polymer, the
CBSi5 must be highly purified prior to attachment to the 1,2-
polybutadiene backbone. The excess TMDS, boiling point (bp)
71 °C, was removed by boiling under vacuum at 50 °C. Since
TMDS is more volatile than the toluene solvent, bp 110 °C,
the product should be completely free of TMDS. Any trace
amounts of water in the system have the potential to form
1,1,3,3,5,5,7,7-octamethyltetrasiloxane (OMTS), which would
not be removed by boiling. Therefore a silica gel column was
used to help remove this and other byproducts of the reaction.
Since water can react with the SiH group and decrease yield,
the column was dried by blowing argon through the column
while heating it with a propane torch. The column was then
loaded with anhydrous hexane as the solvent. After a solution
of toluene/hexane containing the mesogen was added to the
column, a generous amount of hexane was run through the
column to remove the OMTS. Since the mesogen barely moves
down the column with hexane, an anhydrous solution of 33%
toluene in hexane was used to remove it from the column and
to fractionate out other byproducts.
Once purified, the functionalized mesogen CBSi5 was at-
tached to the 1,2-polybutadiene backbone by mixing them in
an anhydrous solution of THF and toluene with a few drops
of the catalyst PCO85. To account for any losses due to
purification, the functionalized mesogen was added at 3
times excess based on the amount of CBV5 initially used
relative to pendant vinyl groups on the polymer backbone. The
excess mesogen also had the effect of pushing the reaction
closer to completion in a reasonable amount of time. The
reaction was allowed to progress for 3 or 4 days even though
80-90% attachment of the mesogen was typically achieved
in 2 days. The reaction was quenched by the addition of an
excess of a vinyl-containing compound, such as styrene, to
neutralize the reactive end of any disiloxane compounds that
may have been partially attached to the polymer backbone.
After a day of quenching the reaction, the product polymer
was precipitated two-three times by addition of methanol to
a solution of the polymer in THF resulting in an overall yield
of 90%.
The fractional attachment of the mesogen was determined
by NMR (Figure 2) and found to be between 87 and 94%.
The polydispersity index (PDI), characterized using gel per-
meation chromatography in THF solution (Waters 401 dif-
ferential refractometer with two 30 mm long Polymer Labo-
ratories PLgel 10 mm analytical columns connected in series,
calibrated against monodisperse polystyrene standards), was
between 1.1 and 1.8. Most of the broadening (the initial
1,2-polybutadienes had a PDI ) 1.04) of the molecular
weight distribution was apparent after the sample was pre-
cipitated out of solution and allowed to stand for a period of a
day following the end of the reaction. After several days the
polymer exhibited long-term stability with virtually no
change in the PDI. It was stored with ∼1% inhibitor, octadecyl
3-(3,5-di-tert-butyl-4-hydroxphenyl)propionate. It is presumed
that broadening of the molecular weight distribution is a
consequence of some polymer-polymer coupling (i.e., cross
linking) due to disiloxanes formed from the dihydroxy-biphenyl
precursor.
This synthesis protocol gave rise to a polymer, designated
PBSiCB5, that was smectic at temperatures below 60 °C and
isotropic at temperatures above roughly 70 °C. Further
characterization was performed by differential scanning cal-
orimetry (DSC; Perkin-Elmer DSC 7 using Pyris software) and
polarizing optical microscopy (POM; Zeiss polarizing micro-
scope with Mettler FP82 hot stage). DSC experiments were
first performed by heating the sample well into the isotropic
phase to remove any thermal history and then cooling at 10
°C/min to a low temperature. Subsequent heating scans at 10
°C/min reveal the polymers to be glassy, with a glass transition
temperature of approximately -10 °C for all samples, and a

6948 Auad et al.
Macromolecules, Vol. 38, No. 16, 2005
Table 1. Physical Property Characterization and Phase
Transition Temperatures for SGLCPs Used in This
Investigation
b
M (kg/mol)a
Mw/Mn
% substitutionc
Ti (°C)d
91
434
1.10
1.10
1.59
1.83
1.90
91
91
94
87
83
79.5
70.0
65.3
63.0
64.0
518
690
1900
a The molecular mass (M) refers to that of the entire side-chain
LCP sample, consisting of both the parent polymer and side group,
and was calculated based on the fractional substitution of me-
sogenic groups. b Polydispersities were measured using GPC. c The
percent substitution was determined by NMR by comparing the
integrated areas of the peaks due to pendant vinyl groups to peaks
associated with the mesogenic groups. d The isotropization tem-
perature (Ti) was detected by DSC and measured at the peak of
the transition.
clearing temperature that varies somewhat among the differ-
ent polymers studied. POM during heating (1-5 °C) shows a
transition from a birefringent liquid-crystal phase below the
DSC-measured transition temperature to an isotropic phase
at high temperatures; however, the mesophase defect texture
was very fine so that no conclusions could be drawn from POM
about the type of liquid-crystal ordering present at lower
temperatures (X-ray diffraction results presented below pro-
vide positive evidence of smectic ordering). Physical properties
and characterization (molecular masses, polydispersities, and
DSC-determined clearing temperatures) of the synthesized
polymers are summarized in Table 1.
Figure 3. Unshifted (a) storage modulus and (b) loss modulus
data for 518 kg/mol PBSiCB5 at indicated temperatures.
2.2. Rheology. Mechanical characterization was performed
using a Rheometrics (RSA II) dynamic mechanical testing
system. Oscillatory strain frequencies typically covered a range
of four decades (0.01-100 rad/s). Temperature control was
achieved by circulating air or nitrogen through the environ-
mental chamber. The uniformity and stability of the temper-
ature were specified to be (0.5 °C. A “shear sandwich”
geometry was used to perform rheological measurements in
which a gap thickness between 0.32 and 0.55 mm was used
for the different samples. Melt films of excess initial thickness
were pressed into the fixtures and subsequently trimmed to
the fixture dimensions. To provide a reproducible initial
condition, samples were heated at least 10 °C into the isotropic
phase and held there for at least 10 min before each set of
experiments to erase effects of sample loading or any previous
strain history. Strain sweep experiments were used to define
the range of linear dynamic response; all frequency sweep data
reported in this paper were collected at small strain amplitude
(<5%) in the linear regime. The effect of large amplitude
oscillatory shear on the structure and rheology of these
polymers is described separately.21
3. Results and Discussion
3.1. Linear Viscoelastic Response and Phase
Behavior of 518 kg/mol PBSiCB5. The storage and
loss moduli, G′ and G′′, (parts a and b of Figure 3,
respectively) of the SGLCPs were measured at temper-
atures ranging from 30 to 80 °C. At the highest
temperatures (T g 70 °C for the 518 kg/mol sample of
Figure 3) G′ and G′′ tend toward conventional terminal
relaxation (G′ ≈ ω2 and G′′ ≈ ω) at low frequencies and
Rouselike behavior (G′, G′′ ≈ ω1/2) at high frequencies.
The failure to perfectly achieve the limiting terminal
slopes may be related to force limitations in these low-
frequency measurements. Within this high-temperature
regime, identified as the isotropic phase using polarized
optical microscopy, the relaxation time decreases with
increasing temperature, while the shapes of the curves
do not change. At somewhat lower temperatures (62.5
< T < 70 °C), the complex modulus changes rapidly,
indicating both a dramatic slowing down of the dynam-
ics and a change of shape in the G′ and G′′ curves, which
no longer show terminal behavior. Finally, at low
temperatures (T e 62.5 °C) the temperature dependence
decreases and the G′ and G′′ curves again adopt self-
similar (but nonterminal) shapes. The general trends
in Figure 3 resemble unshifted linear viscoelastic data
reported by Lee and Han for other smectic SGLCPs.20
Such dramatic changes in the linear viscoelastic re-
sponse are a common signature of ordering transitions.
X-ray scattering was used to elucidate the structural
origins of the transition region between 62.5 and 70 °C
in 518 kg/mol PBSiCB5 (Figure 4). At a temperature of
60 °C, X-ray scattering showed the sample to be smectic,
with a sharp layer reflection peak at low angles (q* )
0.234 Å-1) corresponding to a d spacing of 27 Å indicat-
ing that the side groups are highly interdigitated. When
quenched from the isotropic phase at high temperatures,
there is no macroscopic orientation, and 2-D scattering
patterns are isotropic. However, large amplitude oscil-
2.3. X-ray Experiments. A modified Linkam CSS-450
high-temperature shearing stage, described by Ugaz and
Burghardt,29 was used to carry out small-angle X-ray scatter-
ing (SAXS) experiments. The shear stage provided a parallel
disk shearing geometry with a sample thickness of 1 mm. Thin
Kapton sheets glued to both the fixed and the rotating plates
served as the X-ray windows. An aperture hole located in the
fixed plate defined the size of the incident beam (1 mm
diameter). X-ray experiments were performed on beamline
5BM-D of DND-CAT at the Advanced Photon Source of
Argonne National Laboratory. The highly collimated X-ray
beam had an energy of 25 keV (λ ) 0.496 Å). The use of
relatively high energy confined scattering to smaller angles
allowing both wide- and small-angle scattering to more easily
exit the shear cell. Digital two-dimensional scattering patterns
were collected using a charge-coupled device detector (MarC-
CD). Since a rotating disk geometry was used, scattering
patterns reflect flow-induced variations in molecular orienta-
tion within the plane defined by the flow and vorticity
directions (1,3-plane). Here we only present data collected
under quiescent conditions on previously aligned specimens.

Macromolecules, Vol. 38, No. 16, 2005
Side-Group Liquid-Crystalline Polymers 6949
Figure 6. Effect of temperature on dynamics of 518 kg/mol
PBSiCB5. One over the angular frequency at which G′ )
200 000 dyn/cm2 (see dashed line in Figure 3a) is plotted as a
function of inverse temperature (circles). An overlay compares
these results with the X-ray peak intensity ratio data from
Figure 5 (squares).
cell windows, which provides a convenient internal
standard (Figure 5 insert). This quantifies the gradual
loss of the smectic phase and demonstrates that the
rapid change in dynamics in the temperature range of
62.5-70 °C is associated with a broad smectic-isotropic
biphasic region.
Figure 4. X-ray scattering patterns for 518 kg/mol PBSiCB5.
(a) Quiescent 2-D pattern in the smectic phase (60 °C) after
sample had been sheared at 0.05 Hz and 95% strain for 62
min, followed by relaxation for 12 min. The isotropic ring at
intermediate q arises from the Kapton windows in the shear
cell. Also shown are quiescent images expanded to emphasize
changes in smectic reflection during heating at 61 (b), 65 (c),
69 (d), and 73 °C (e).
A simple quantitative analysis of the rheological data
allows determination of the width of the biphase region.
We take one over the angular frequency at which G′ )
200 000 dyn/cm2 as a convenient indicator of the changes
in time scale with frequency over the temperature range
of interest (an arbitrary choice of condition, but one
which is encompassed by most of the data sets in Figure
3). These results show that the temperature dependence
clearly breaks into three regions and, further, that the
intermediate regime of rapid change is well correlated
with the X-ray results indicating a smectic-isotropic
biphase (Figure 6). Intersections between lines fit to
data in each regime allows accurate rheological deter-
mination of the location and width of the biphasic
regime. A similar procedure was used for samples of
varying molecular weight; in all samples studied using
both rheology and X-ray scattering, good agreement was
found with these independent measures of the location
and width of the biphasic regime (Table 2). We attribute
the biphasic region to heterogeneity in the degree of
substitution of the mesogenic side chains. This is
discussed further in section 3.3 below.
3.2. Temperature Dependence of Dynamics in
518 kg/mol PBSiCB5. A Cole-Cole plot (G′ vs G′′,
Figure 7) demonstrates thermo-rheological complexity
as this material passes through its ordering transition;
the failure of data to superimpose indicates differences
in the shape of the relaxation spectrum in the various
temperature regimes. At the same time, data collected
within either the smectic or isotropic phases do sepa-
rately collapse in the Cole-Cole representation, indicat-
ing temperature independence in the shape of the
relaxation spectrum. This provides justification for
using time-temperature shifting to explore the tem-
perature dependence of dynamics in either phase. While
the ability to shift data within the isotropic phase is not
surprising, there are many reasons to expect that this
might not be possible within the smectic phase. Tem-
perature-dependent changes in the thermodynamic
state of the SGLCP (e.g., order parameter) or contribu-
tions of distortional elasticity to the measured moduli
are two examples of phenomena that could lead to more
Figure 5. Azimuthally averaged X-ray scattered intensity,
I(q), for 518 kg/mol PBSiCB5 at different temperatures. Insert
shows ratio of intensity of smectic to Kapton peak as a function
of temperature.
latory shear is capable of inducing macroscopic align-
ment in the sample (Figure 4a). The low q diffraction
from smectic layers is concentrated along the vorticity
direction, indicating that shear promotes orientation in
the “perpendicular” state, where layers lie within the
flow-gradient plane. The diffuse wide-angle mesogen
peak at higher q is also rendered anisotropic by shear,
with intensity concentrated perpendicular to the smectic
reflections, indicating ordering of the smectic-A type.
The anisotropic, shear-aligned PBSiCB5 sample was
quiescently heated in 1 °C increments through the range
of temperatures (60-80 °C) where the dramatic changes
in rheology are observed. Upon heating from the smectic
state, a broad smectic-isotropic transition is manifested
by the gradual reduction in intensity and broadening
of the low q smectic reflection (parts b-e of Figure 4).
Azimuthally averaged I(q) scans in the small-angle
regime show a systematic drop in smectic peak intensity
with increasing temperature (Figure 5). To quantify
these changes, we take the ratio of the smectic peak
height to that of the peak arising from the Kapton shear

6950 Auad et al.
Macromolecules, Vol. 38, No. 16, 2005
Table 2. Phase Transition Temperatures Determined through Rheology and X-ray Scattering Experiments
M (kg/mol)
Ts-biph. (°C)a
Tbiph.-i (°C)a
∆Ti-s (°C)
Ts-biph. (°C)b
Tbiph.-i (°C)b
∆Ti-s (°C)
91
434
82.5
75.0
70.0
65.0
63.5
55.0
62.5
61.0
61.0
20
7.5
4
75.0
70.0
65.0
63.0
518
62.0
63.0
62.0
8
2
1
690
1900
2
a Phase transition temperatures determined rheologically. b Transition temperatures determined by X-ray scattering.
Figure 7. Cole-Cole plot of the storage modulus vs the loss
modulus for 518 kg/mol PBSiCB5 at different temperatures.
complex temperature-dependent changes in the relax-
ation spectrum. Despite these potential complications,
the data themselves testify that superposition is possible
in this case. Indeed, previous literature results have
generally demonstrated the ability to construct master
curves within smectic phases via time-temperature
shifting,6,13,19 and this procedure has been widely adopted.
Data collected in the smectic (T e 62.5 °C) and
isotropic (T g 70 °C) phases were first shifted to create
separate master curves. Over the temperature ranges
studied, the temperature dependence could be described
by an Arrhenius expression with activation energies of
45 and 61.5 kJ/mol, respectively, for the isotropic and
smectic phases. To clarify the nature of the rheological
changes associated with the phase change itself, these
master curves were shifted to a common temperature
of 67.5 °C, within the biphasic region, via small extrapo-
lations using the activation energies measured within
either phase (Figure 8). This procedure is similar to that
adopted by Colby et al. in which separate master curves
from smectic and nematic phases of a SGLCP were
shifted to a common reference temperature.6 As in that
case, this representation reinforces the dramatic quali-
tative and quantitative changes to the dynamics associ-
ated with smectic ordering (although here, of course, it
is the smectic and isotropic phases that are being
compared). Data collected within the biphasic region at
T ) 67.5 °C are surprisingly well represented by a
simple linear mixing rule
Figure 8. (a) Storage and (b) loss modulus of 518 kg/mol
PBSiCB5, in which smectic and isotropic data have been
separately time-temperature superimposed and then shifted
to 67.5 °C (a temperature in the biphasic region) by small
extrapolations using the activation energies measured in each
phase. Solid curves represent generalized Maxwell fits to the
smectic and isotropic data, while the heavy dashed curve
represents prediction of a linear mixing rule, eq 1, using x )
0.13. Arrows represent a possible interpretation of the rheo-
logical changes accompanying the isotropic-smectic transition;
see discussion in text.
ditional ordered fluid structure or, alternatively, to a
slowing down of the chain dynamics caused by the
topological restrictions imposed by localization of meso-
gens in smectic layers. The latter viewpoint has been
frequently adopted in the literature. Drawing upon
procedures originally developed in the context of mi-
crophase separation in lamellar block copolymers,30
Rubin and co-workers13 used time-temperature shifting
to collapse data collected in the isotropic, nematic, and
smectic phases of different SGLCP polymers in the limit
of high frequency, hypothesizing that the shape of the
relaxation spectrum at high frequencies (i.e., local chain
dynamics) will be unaffected by ordering at larger length
scales. Similar procedures have subsequently been
adopted by other investigators,19 although the use of any
time-temperature shifting procedures in ordered poly-
mers has also received criticism.20 Partial support for
this approach in the present case comes from the near
superposition of the high-frequency isotropic data with
the smectic data in the Cole-Cole plot (Figure 7).
Gh * ) xGi* + (1 - x)Gs*
(1)
which interpolates between the smectic and isotropic
results as x increases from 0 to 1. (Predictions of eq 1
in Figure 8 were generated by first fitting the smectic
and isotropic relaxation moduli to a generalized Max-
well spectrum and then using these analytical expres-
sions to compute Gh *.) From Figure 3, it is clear that
decreasing temperature within the biphase leads to
rheology closer to that found in the smectic phase, as a
larger fraction of the polymer undergoes ordering.
The changes in G′ and G′′ upon ordering may be
attributed to an increase in modulus due to the ad-

Macromolecules, Vol. 38, No. 16, 2005
Side-Group Liquid-Crystalline Polymers 6951
Results of a shifting procedure which attempts su-
perposition of all data at high frequency are presented
in Supporting Information available for this paper.
These results allow comparison to other literature
reports in which this procedure has been used,13,19 but
we do not believe the procedure is warranted in this
case. First and foremost, it is not obvious that concepts
applied successfully to lamellar block copolymers should
necessarily translate to SGLCPs. In the former case, the
ordered layers possess a length scale comparable to the
radius of gyration of the full chain, so that a reasonable
separation of time scales for layer dynamics and the
Rouselike dynamics of subchains may be expected. For
entangled hydrocarbon block copolymers of higher mo-
lecular weight, the entanglement length provides an
additional intermediate length scale whose dynamics
are still observable within the ordered lamellar phase.30
Conversely, the ordered layers in smectic SGLCPs adopt
the length scale of the mesogens and hence the mono-
mers themselves. Under conditions in which a single
statistical segment of the chain in the isotropic melt may
possibly participate in multiple smectic layers upon
ordering, it is difficult to imagine that any length scale
is sufficiently “local” (except perhaps at the very high
frequencies of the dynamic glass transition) to render
high-frequency chain dynamics in the isotropic and
smectic phases similar enough to rationalize this type
of shifting.
Rubin and co-workers did see clear rheological conse-
quences of nematic ordering at low frequency13 but were
using very high molecular weight polymers relative to
that studied by Lee and Han.)
In the absence of isotropic phase data at much higher
frequencies (a difficult proposition for routine methods
of melt rheometry), we favor the representation of
Figure 8 for the SGLCPs considered here. This presen-
tation highlights the rheological consequences of meso-
phase formation without imposing an implicit assump-
tion that these rheological changes only reflect a slowing
down of dynamics (i.e., a horizontal shift). While it is
probably impossible to decompose these changes with
certainty, it is reasonable to postulate smectic ordering
of SGLCPs might be accompanied by both a slowing
down of dynamics and an additional contribution to the
modulus, which would move the curves vertically. In the
case of the PBSiCB5 sample in Figure 8, it is interesting
to note that the smectic phase data are not a featureless
power law but show signs of a relaxation process at
intermediate frequencies. If one postulates that this is
related to the overall chain dynamics (which we recog-
nize to be a speculative assumption), comparison with
the chain relaxation in the isotropic data suggests that
smectic ordering leads to both dynamic slowing (hori-
zontal shift) as well as modulus enhancement (vertical
shift), as represented by the arrows in Figure 8. This
notion is similar in spirit to the observations of Colby
et al. that enhanced elastic character in a smectic
SGLCP extends to much lower frequencies than esti-
mates of the time scale for overall chain motion from
diffusion measurements, suggesting that it is associated
with structural elements much larger than the typical
coil size (e.g., smectic grain structures).6
3.3. Effect of Fractional Mesogenic Substitution
and Molecular Weight on the Biphasic Regime. A
series of polymers having the same mesogenic structure
but differing in their molecular weight were made in
order to determine the effect on polymer rheology and
phase behavior. As shown in Table 2, all the samples
had a biphasic region between the isotropic and smectic
phases. Maintaining low polydispersity while achieving
a high degree of attachment of mesogenic groups was
difficult at larger molecular weights; therefore, an
increase in molecular weight was accompanied by an
increase in polydispersity and a decrease in the degree
of attachment. Tables 1 and 2 demonstrate that higher
molecular samples generally showed lower transition
temperatures and narrower biphasic windows. Which
molecular parameters control the temperature and
width of the smectic-isotropic transition? Wewerka et
al. found that a higher degree of polymerization of the
polymer backbone increased isotropic-to-mesophase tran-
sition temperatures in their side-chain polymers,19 while
the opposite trend is observed here. Their work consid-
ered molecular weights that were much lower than
those of our samples, such that chain-end effects could
play a significant role. Since a chain end is likely to be
a source of local disruption in the liquid-crystalline
order, a greater fraction of chain ends leads to a greater
perturbation to the LC structure and hence to a lower
clearing temperature. Conversely, our study employed
very large molecular weight samples, where effects of
chain ends on transition temperatures should be neg-
ligible. Under these conditions, molecular weight poly-
dispersity is also not expected to lead to significant
broadening of the transition; for instance, Kasko et al.31
The coincidence of smectic and isotropic data in a
Cole-Cole plot such as Figure 7 may, indeed, be mere
coincidence. As in the case of the 518 kg/mol PBSiCB5
polymer highlighted here, such collapse of the data often
involves high-frequency isotropic data in which Rouse-
like behavior, G′, G′′ ≈ ω1/2, is observed (as befits the
notion of local chain motion). However, ordered fluids
with layered morphologies frequently also exhibit the
same power-law scaling.18,30 By the Kramers-Kronig
1
relations, power-law rheology with an exponent of /2
must also be characterized by G′ ) G′′. In other words,
Rouselike and typical layerlike rheological behavior
must fall along a common line on a Cole-Cole plot.
Superposition of the type seen in Figure 7 does not
necessarily signify similarity of mechanism at the
molecular or mesoscopic level.
Attempts to create master curves in these smectic
SGLCPs by superposing “high” frequency data are also
limited by the narrow range of overlap between data
sets, itself a consequence of the dramatic rheological
changes observed upon ordering (Figure 3). In the
earlier work of Rubin and co-workers, the dynamic
differences between the isotropic, nematic, and smectic
phases were less dramatic than those found here,
leading to a wider range of overlap in the master curves
obtained by this procedure. Certainly, the superposition
of isotropic and nematic data by Rubin and co-workers
seems robust given the wide range of frequency overlap
available and the high quality of the superposition
achieved at high frequencies;13 also note that Colby et
al. successfully superimposed data from isotropic and
nematic phases.6 In this vein, it is interesting to note
that Lee and Han observed a nearly perfect superpo-
sition of data for isotropic and nematic phases in a
G′-G′′ plot for a nematic SGLCP, from which they
conclude that the dynamic consequences of the isotropic-
nematic transition are not as severe as for the isotropic-
smectic transition,20 a conclusion consistent with earlier
observations of nematic SGLCP rheology.6-8 (Recall that

6952 Auad et al.
Macromolecules, Vol. 38, No. 16, 2005
Figure 9. Effect of PBSiCB5 molecular weight on the linear viscoelastic response. Rheological master curves for (a) 434, (b) 690,
and (c) 1900 kg/mol PBSiCB5 samples were prepared using the procedures of Figure 7, in which separate smectic and isotropic
phase master curves were prepared and shifted to a common reference temperature in the biphase region for each sample. Dashed
curves represent the prediction of a linear mixing rule, eq 1, using (a) x ) 0.12, (b) x ) 0.40, and (c) x ) 0.56.
found that polymers having a broad Gaussian-like
distribution of chain lengths (i.e., broader PDI) did not
exhibit broadening of the Ts-i region. Instead, the
changes in Tbiph.-i and ∆Ti-s observed here are most
likely the result of changes in the degree of mesogenic
substitution rather than the DP or polydispersity.
it is statistically more likely that each molecule in a
random copolymer will approach the mean composition
of the sample as a whole. This reduction in the compo-
sitional heterogeneity of the sample would result in a
corresponding decrease in the width of the two-phase
region. Detailed testing of these hypotheses would
require further experiments on samples with varying
molecular weights but the same degree of substitution
(and vice versa), which presents a significant challenge
in synthesis and characterization.
3.4. Effect of Molecular Weight on Rheology.
Smectic ordering leads to similar rheological conse-
quences for all molecular weights of PBSiCB5 studied
(Figure 9). Comparison of isotropic and smectic data
shifted to a common temperature in the biphase reveals
evidence of retarded dynamics and/or moduli enhance-
ment for all molecular weights studied. In each case,
the simple linear mixing rule (eq 1) can provide a
reasonable representation of data collected in the bi-
phase region. Note that because of the sample-depend-
ent location and width of the biphase, and the particular
temperatures chosen for rheological study, these plots
differ slightly in the choice of reference temperature
within the biphase regime.
Similar to the role of chain ends in lower molecular
weight samples, monomeric units without mesogens
would be expected to disrupt molecular packing and
consequently lower the transition temperature. Any
branches associated with sparse cross linking in the
higher molecular weight samples (the presumed origin
of the higher polydispersity) would also serve to disrupt
molecular packing and suppress Ts-i; however, given the
reduced degree of mesogen substitution found at high
molecular weights (Table 1), “missing mesogen defects”
should be much more prevalent than branch points and
are thus likely the primary factor explaining the trend
toward lower clearing points. The consequences of
random and incomplete mesogen attachment can be
considered in relation to concepts of Stupp et al.,32 who
discuss the origins of a biphasic region in a melt of a
main chain nematic random copolymer of monomers of
different intrinsic stiffness. They invoke the concept of
“polyflexibility,” referring to a distribution of chain
flexibility, reflecting in turn the statistical distribution
of composition among individual molecules in a random
copolymer. This varying degree of flexibility will cause
the stiffer molecules to transition at lower temperatures
than more flexible molecules, leading to a biphasic
region. This reasoning transfers to our systems in that
incomplete mesogen attachment renders our SGLCPs
random copolymers. Individual molecules within the
sample will thus exhibit a statistical distribution of
degree of mesogen attachment, allowing a biphasic
region in which molecules with higher mesogen content
are preferentially partitioned into the smectic phase.
As expected, the isotropic phase data show a strong
dependence on molecular weight and, in particular,
demonstrate the emergence of an entanglement plateau
for the higher molecular weight samples. The plateau
modulus is estimated as the value of G′ where the phase
angle is at its minimum,33 leading to GN ≈ 1.15 × 105
0
dyn/cm2, quite similar to that found in the SGLCP
studied by Rubin and co-workers (1.2 × 105 dyn/cm2).13
Conversely, there is only a weak effect of molecular
weight on the rheology within the smectic phase,
consistent with the hypothesis that the smectic rheology
is dominated by layer dynamics, in which polymer
molecular weight does not play a significant role. There
do appear to be small qualitative changes in the low
frequency portion of the smectic response as molecular
weight increases and entanglement dynamics emerge.
These concepts may be extended to rationalize why
the width of the biphasic region decreases in samples
of larger molecular weight (Table 2). As DP increases,

Macromolecules, Vol. 38, No. 16, 2005
Side-Group Liquid-Crystalline Polymers 6953
Supporting Information Available: Text discussing a
shifting procedure within a single master plot and figures
showing storage modulus and loss modulus master curves
for 518 kg/mol PBSiCB5 and time-temperature shift fac-
tors obtained from shifting the data from Figure 8. This
material is available free of charge via the Internet at
http://pubs.acs.org.
Higher molecular weights (in which the plateau extends
to lower frequencies) would presumably lead to more
dramatic changes in the low-frequency smectic response.
4. Conclusions
We have synthesized new model 1,2-polybutadiene-
based thermotropic SGLCPs with cyano-biphenyl meso-
genic side groups with relatively low polydispersity and
high molecular weight. Rheology and X-ray methods
were used to characterize their phase structure and
viscoelastic properties and showed that these polymers
exhibit smectic ordering in the liquid crystalline state.
The linear viscoelastic behavior is strongly affected by
the smectic ordering. The temperature dependence of
the dynamics reveals the existence of three distinct
regimes, which X-ray scattering demonstrated to be
smectic at low temperatures (T < 60 °C), isotropic at
high temperatures (T > 80 °C), and an intermediate
smectic-isotropic biphase. The presence of this biphasic
region is attributed to the incomplete attachment of the
mesogenic units to the pendant vinyl groups of the 1,2-
polybutadiene pre-polymer and the resulting statistical
distribution of composition within the polymer sample.
This concept allows rationalization of how the location
and width of the biphase region varies with polymer
composition and molecular weight. While smectic SGLCP
data have frequently been shifted to create master plots
in which isotropic and smectic data are superimposed
in the high-frequency limit,13,19 we question whether
such a representation can be justified in the present
case. We suggest an alternate representation of the data
that emphasizes rheological consequences of ordering
while requiring fewer assumptions. The isotropic phase
rheology is strongly dependent on molecular weight and
shows the emergence of an entanglement plateau for
high molecular weight samples. Conversely, the smectic
phase rheology does not depend as strongly on polymer
molecular weight, consistent with the assumption that
the viscoelastic response is dominated by the smectic
layers.
References and Notes
(1) Hsu, C. S. Prog. Polym. Sci. 1997, 22, 829.
(2) Shibaev, V. P.; Kostromin, S. G.; Plate, N. A.; Ivanov, S. A.;
Vetrov, V. Y.; Yakovlev, I. A. Polym. Commun. 1983, 24, 364.
(3) Eich, M.; Wendorff, J. H.; Ringsdorf, H.; Schmidt, H. W.
Makromol. Chem. 1985, 186, 2639.
(4) Cabrera, I.; Krongauz, V. Macromolecules 1987, 20, 2713.
(5) Cabrera, I.; Krongauz, V.; Ringsdorf, H. Angew. Chem., Int.
Ed. 1987, 26, 1178.
(6) Colby, R. H.; Gillmor, J. R.; Galli, G.; Laus, M.; Ober, C. K.;
Hall, E. Liq. Cryst. 1993, 13, 233.
(7) Kresse, H.; Stettin, H.; Tennstedt, E.; Kostromin, S. Mol.
Cryst. Liq. Cryst. 1990, 191, 135.
(8) Findlay, R. B. Mol. Cryst. Liq. Cryst. 1993, 231, 137.
(9) Porter, R. S.; Johnson, J. F. In Rheology; Eirich, F. R., Ed.;
Academic Press: New York, 1967; Vol. 4, Chapter 5.
(10) Wissbrun, K. F.; Griffin, A. C. J. Polym. Sci. Pol. Phys. 1982,
20, 1835.
(11) Baird, D. G. In Liquid Crystalline Order in Polymers;
Blumstein, A. Y., Ed.; Academic Press: New York, 1978.
(12) Papkov, S. P.; Kulichik, Vg; Kalmykov, Vd; Malkin, A. Y. J.
Polym. Sci. Pol. Phys. 1974, 12, 1753.
(13) Rubin, S. F.; Kannan, R. M.; Kornfield, J. A.; Boeffel, C.
Macromolecules 1995, 28, 3521.
(14) Berghausen, J.; Fuchs, J.; Richtering, W. Macromolecules
1997, 30, 7574.
(15) Fourmaux-Demange, V.; Brulet, A.; Cotton, J. P.; Hilliou, L.;
Martinoty, P.; Keller, P.; Boue´, F. Macromolecules 1998, 31,
7445.
(16) Quijada-Garrido, I.; Siebert, H.; Friedrich, C.; Schmidt, C.
Macromolecules 2000, 33, 3844.
(17) Lee, K. M.; Han, C. D. Macromolecules 2002, 35, 6263.
(18) Larson, R. G.; Winey, K. I.; Patel, S. S.; Watanabe, H.;
Bruinsma, R. Rheol. Acta 1993, 32, 245.
(19) Wewerka, A.; Viertler, K.; Vlassopoulos, D.; Stelzer, F. Rheol.
Acta 2001, 40, 416.
(20) Lee, K. M.; Han, C. D. Macromolecules 2003, 36, 8796.
(21) Rendon, S.; Burghardt, W. R.; Auad, M. L.; Kornfield, J. A.,
in preparation.
(22) Halasa, A. F.; Lohr, D. F.; Hall, J. E. J. Polym. Sci., Part A:
Acknowledgment. We gratefully acknowledge fi-
nancial support from the Air Force Office of Scientific
Research (AFOSR)-LC MURI (f4962-97-1-0014), Fun-
dacio´n Antorchas (Argentina), CONICET (Argentina),
and the National Science Foundation (DMI-0132519).
X-ray scattering experiments were conducted at the
DuPont-Northwestern-Dow Collaborative Access Team
(DND-CAT) Synchrotron Research Center located at
Sector 5 of the Advanced Photon Source of Argonne
National Laboratory. DND-CAT is supported by the E.I.
DuPont de Nemours & Co., the Dow Chemical Com-
pany, and the National Science Foundation through
Grant DMR-9304725 and the State of Illinois through
the Department of Commerce and the Board of Higher
Education Grant IBHE HECA NWU 96. Use of the
Advanced Photon Source was supported by the U.S.
Department of Energy, Basic Energy Sciences, Office
of Energy Research, under Contract No. W-31-102-Eng-
38.
Polym. Chem. 1981, 19, 1357.
(23) Bywater, S.; MacKerron, D. H.; Worsfold, D. J.; Schue, F. J.
Polym. Sci., Polym. Chem. Ed. 1985, 23, 1997.
(24) Sanger, J.; Tefehne, C.; Lay, R.; Gronski, W. Polym. Bull.
1996, 36, 19.
(25) Hempenius, M. A.; Lammertink, R. G. H.; Vancso, G. J.
Macromol. Rapid Commun. 1996, 17, 299.
(26) Hempenius, M. A.; Lammertink, R. G. H.; Vancso, G. J.
Macromolecules 1997, 30, 266.
(27) Moment, A.; Miranda, R.; Hammond, P. T. Macromol. Rapid
Commun. 1998, 19, 573.
(28) Moment, A.; Hammond, P. T. Polymer 2001, 42, 6945.
(29) Ugaz, V. M.; Burghardt, W. R. Macromolecules 1998, 31,
8474.
(30) Rosedale, J. H.; Bates, F. S. Macromolecules 1990, 23, 2329.
(31) Kasko, A. M.; Grunwald, S. R.; Pugh, C. Macromolecules
2002, 35, 5466.
(32) Stupp, S. I.; Wu, J. L.; Moore, J. S.; Martin, P. G. Macro-
molecules 1991, 24, 6399.
(33) Wu, S. J. Polym. Sci., Polym. Phys. Ed. 1987, 25, 2511.
MA050551F