Tailored phase transitions via mixed-mesogen liquid crystalline polymers with silicon-based spacers

Article abstract of DOI:10.1021/ma048327y

Control over thermotropic phase behavior in low-T_g main-chain liquid crystalline polymers (LCPs) is desired for a variety of applications, including soft actuation when cross-linked. Here, we describe the synthesis of new silicon-based main-cha

Full text of DOI:10.1021/ma048327y

Macromolecules 2005, 38, 4103-4113
4103
Tailored Phase Transitions via Mixed-Mesogen Liquid Crystalline
Polymers with Silicon-Based Spacers
Ingrid A. Rousseau,†,‡ Haihu Qin,†,§ and Patrick T. Mather*,†,|
Institute of Materials Science and Chemical Engineering Department, University of Connecticut,
Storrs, Connecticut 06269
Received August 13, 2004; Revised Manuscript Received February 22, 2005
ABSTRACT: Control over thermotropic phase behavior in low-Tg main-chain liquid crystalline polymers
(LCPs) is desired for a variety of applications, including soft actuation when cross-linked. Here, we describe
the synthesis of new silicon-based main-chain LCPs, including homopolymers, blends, and copolymers,
with tunable clearing temperatures as governed by their chemical composition. Two mesogenic groups,
namely, 1,4-bis[4-(4-pentenyloxy)benzoyl]hydroquinone (M1) and 2-tert-butyl-1,4-bis[4-(4-pentenyloxy)-
benzoyl]hydroquinone (M2), were polymerized with various silicon-based flexible spacers, specifically,
1,4-bis(dimethylsilyl)benzene (S1), 1,1,3,3,5,5-hexamethyltrisiloxane (S2), and hydride-terminated poly-
(dimethylsiloxane) (DP ) 8) (S3) spacers, following routine hydrosilation reaction techniques. These
mesogens and flexible spacers were chosen so that both copolymerization and blending of homopolymers
would allow for potential tailoring of phase behavior. Indeed, despite their similar chemical structure,
the clearing transition temperatures of M1 and M2 differ dramatically (∆TNI ) 140 °C), while the silicon-
based spacers offer accessibility to a large range of molecular flexibility. High-molecular-weight LCPs
were successfully prepared using Pt-catalyzed addition polymerization. Interestingly, the polymers
exhibited wide liquid crystalline windows with relatively high degree of order (smectic phases) except for
the S1-based blends, which, in addition to a smectic phase, also displayed a narrow nematic phase. As
expected, a drastic decrease of the glass transition temperature arose on polymerizing with longer, more
flexible spacers, from about 56 to -17 °C. Finally, in comparing the two approaches to phase behavior
tailoring, namely, blending vs copolymerization, the former led to apparently immiscible systems with
constant isotropization temperatures, while the latter yielded homogeneous, single-phased materials with
tunable isotropization temperatures dictated by the M1/M2 ratio of the copolymers.
Introduction
hibit comparatively high transition temperatures rela-
tive to room temperature, so that, to date, they have
not received adequate attention. In addition, aside from
some studies examining network architecture variation,
only limited attention has been given to other important
synthetic variables when dealing with cross-linked
structures, particularly the influence of mesophase type,
nematic, cholesteric, or various smectic on the resulting
thermomechanical behavior.
Glassy side-chain liquid crystalline polymers (SC-
LCPs) have been widely studied in the past1 as a
materials approach that uniquely combines the me-
chanical and thermal properties of polymers with the
optical properties of small-molecule liquid crystals, with
numerous possible applications.2 It is over the past two
decades that researchers have focused more on cross-
linked SC-LCPs because of the interesting thermome-
chanical properties they exhibit as liquid crystalline
elastomers (LCEs).3 Indeed, side-chain nematic LCEs
have been shown to display spontaneously large strain-
reversible actuation and soft elasticity when exposed to
specific stimuli.4,5 Here, a thermally stimulated actua-
tion behavior, shrinking on heating through a clearing
transition and expanding on cooling through the same,
has been explained by a coupling between liquid crys-
talline order and rubber elasticity resulting from the
underlying cross-linked structure.6 Yet higher actuator
performance has been anticipated7 and recently dem-
onstrated for main-chain liquid crystalline elastomers
(MC-LCEs) because of an enhanced coupling between
their intrinsically high, yet labile, orientational order
and network strain compared to their side-chain ana-
logues.8 Challenges exist, however, for main-chain liquid
crystalline polymers, particularly regarding their syn-
thesis and subsequent processing into elastomers. More
specifically, main-chain LCPs (MC-LCPs) usually ex-
A general lack of knowledge regarding the structure-
property relationships in these materials, coupled with
their high potential to yield improved thermomechanical
properties (i.e., soft actuation), has led us to study new
MC-LCPs for eventual incorporation into elastomeric
structures by chemical or physical cross-linking. We are
particularly interested in understanding the influence
of mesogen structure and flexible spacer length as well
as the various polymer architectures, homopolymers,
blends, and copolymers, on liquid crystalline phase
behavior and glass transition temperature. Such an
understanding will allow tailoring of associated ther-
momechanical (actuation) behavior in related cross-
linked structures so that their utility in strongly re-
strictive environments may be facilitated. An example
category of such restrictive applications include bio-
medical devices, where both biocompatibility and low
temperature activation are required. In the area of low
temperature activation, we recently reported the shape-
memory behavior of a new siloxane-based main-chain
smectic-LC elastomer that was shown to allow the fixing
of large strains for subsequent recovery on heating
through the smectic-C-to-isotropic transition.9
* Author to whom correspondence should be addressed. E-
mail: patrick.mather@case.edu.
† Institute of Materials Science.
‡ ingrid@mail.ims.uconn.edu.
§ haihu.qin@case.edu.
Here, we report on a parallel effort in developing
structure-property relationships in analogous linear
| Chemical Engineering Department.
10.1021/ma048327y CCC: $30.25 © 2005 American Chemical Society
Published on Web 04/22/2005

4104 Rousseau et al.
Macromolecules, Vol. 38, No. 10, 2005
in a flowing N2 atmosphere. Indium was used as a calibration
standard for both the temperature and heat flow scales.
Samples with weights ranging from 5 to 10 mg were encap-
sulated in aluminum pans for testing. Temperatures corre-
sponding to maxima of the DSC endothermic peaks of the
second heating traces were assigned as phase transition
temperatures. The midpoint of the heat capacity stepwise
increase was taken as the glass transition temperature, Tg,
when appropriate.
polymers bearing various silicon-based spacers so as to
enable property tailoring of similar structures when
cross-linked into network form. We present the synthe-
sis of two chemically similar mesogens, namely, 1,4-bis-
[4-(4-pentenyloxy)benzoyl]hydroquinone10,11 (M1) and
2-tert-butyl-1,4-bis[4-(4-pentenyloxy)benzoyl]hydroquino-
ne (M2), that, although chemically similar, differ drasti-
cally in their phase transition temperatures. Indeed, we
will show that substituting the slender mesogen, M1,
with a tert-butyl group (M2) results in a 140 °C drop in
the nematic-isotropic transition, thus suggesting that
one could tune the thermal properties of the resulting
materials upon mixing of these mesogens; i.e., “mixed-
mesogen” LCPs. We thus describe the synthesis and
characterization of silicon-based main-chain liquid crys-
talline homopolymers, blends, and copolymers using
both M1 and M2 as mesogens. Three different silicon-
based compounds were used as flexible spacers, namely,
1,4-bis(dimethylsilyl)benzene (S1), 1,1,3,3,5,5-hexa-
methyltrisiloxane (S2), and hydride-terminated poly-
(dimethylsiloxane) (DP ) 8) (S3). In addition to the effect
of mesogen structure on phase behavior, we also reveal
the influence of siloxane-based spacer structure on
physical and thermooptical properties of the resulting
polymers.
Polarizing optical microscopy (POM) studies were performed
using an Olympus BX50 microscope equipped with crossed
polarizers, a STC-200 hot stage from Instec Inc., and a
composite color CCD camera (Panasonic GP-KR222). Images
were acquired from the CCD camera at selected times and/or
temperatures using a frame grabber and Linksys software
(Linkam Scientific). Spatial dimensions were calibrated using
a stage micrometer with 10 µm line spacing. Unless otherwise
noted, a 20x/0.4 NA achromat long working-distance objective
lens (Olympus LMPlanFI) was employed. The samples used
for POM analysis were sandwiched between two glass cover-
slips and melted on a hot stage at 150 °C, care being taken to
avoid coverslip flexure that would lead to void formation, and
subsequently cooled to room temperature. The temperature
ramping rates were chosen to be consistent with DSC experi-
ments for comparison purposes, and shown in this paper are
micrographs obtained on second heating of the samples after
an initial melting at 150 or 200 °C followed by a cooling ramp
at 5 or 10 °C/min for the copolymers and blends, respectively.
The hot stage was equipped with a liquid nitrogen LN2-P
cooling unit from Instec, Inc. for accurate control of the sample
temperature, either isothermally or during heating and cooling
runs.
Wide-angle X-ray diffraction (WAXD) experiments were
performed either on fibers drawn from the melt or on powdered
samples using a Bruker AXS instrument with a chromium
source (λ ) 2.291 Å) or a Bruker AXS D8 Advance using a
CuKR source (λ ) 1.5418 Å). The latter was used both at room
temperature and at elevated temperatures, in selected cases,
with the aid of a nickel heating strip and digital temperature
controller. The former was used for stretched specimens and
employed a sample-detector distance of 6 cm. In both cases,
the X-ray power source was operated at 40 mA and 40 kV.
Data were gathered and analyzed via a general area detector
diffraction system (GADDS) software version 3.317 or a
Diffraction Plus software version 5.0.
Experimental Section
Materials and Methods. 1,1,3,3,5,5-hexamethyltrisiloxane
(S2) was purchased from Gelest, Inc., whereas the hydride-
terminated poly(dimethylsiloxane) (S3) was obtained from both
Gelest, Inc. and Aldrich. 1,4-bis(dimethylsilyl)benzene (S1),
tetrakis(vinyldimethylsiloxy)silane (CL), R,ω-divinyl-termi-
nated poly(dimethylsiloxane), and the platinum-based catalyst
(platinum(0)-1,3-divinyl-1,1,3,3-tetramethyldisiloxane complex
in xylenes), were purchased from Aldrich with no information
regarding purity given by the vendor. Hydroquinone (99%) was
purchased from Aldrich and used without further purification.
Anhydrous dichloromethane (99.9%) was purchased from Acros
and dichoromethane (Optima) from Fisher Scientific. All other
solvents were purchased from Acros. All chemicals were used
without further treatment except for tert-butylhydroquinone
(97%), which was purified by recrystallization from toluene
(approximately 20 g/L) to yield white crystals. When required,
analytical thin-layer chromatography (TLC) was conducted
using precoated silica gel 60 F254 plates (E. Merck).
Monomer Synthesis. Scheme 1 represents two different
synthetic routes for the preparation of two mesogenic dienes,
namely, 1,4-bis[4-(4-pentenyloxy)benzoyl] hydroquinone (M1)
and 2-tert-butyl-1,4-bis[4-(4-pentenyloxy)benzoyl] hydroquino-
ne (M2). To the best of our knowledge, M2 is a new compound
and so its synthesis steps are detailed below. The synthesis
and purification of M1 are very close to those of M2 and have
been previously reported by other researchers.10,11 Therefore,
the preparation of M1 is not included in this report.
To verify the chemical structures synthesized, liquid phase
1H NMR characterizations were performed using a Bruker
AVANCE DMX500 spectrometer. The samples were prepared
in either D6-acetone or CDCl3 at room temperature, depending
on individual solubilities and chemical shifts, with tetra-
methylsilane (TMS) added as an internal standard.
Gel permeation chromatography (GPC, Waters Associates,
150-C Plus) with a PL-ELS 1000 evaporative light scattering
detector (ELSD, Polymer Laboratories) was used to obtain
molecular weights (Mh n and Mh w) relative to monodispersed
polystyrene standards (472, 982, 4000, 6930, 43 000, 200 000,
400 000, and 824 000 g/mol; Polymer Standards Service-USA,
Inc.), and polydispersity index (PDI ) Mh w/Mh n). The samples,
dissolved in THF to about 0.1 wt %, were injected at 35 °C
with THF as an eluant and at a flow rate of 1.0 mL/min. A set
of three columns packed with cross-linked divinylbenzene in
series, thermostated at 35 °C, was used. Note that the
resulting values for average molecular weights and polydis-
persity will not be absolute values since the chain conformation
adopted by linear polystyrene (standards) is expected to be
quite different from that of the liquid crystalline polymer
investigated here. However, they will allow for qualitative
comparison of the values obtained for the various LCP series.
Synthesis of 5-Pentenyloxybenzoic Acid (1). In a three-
neck 500-mL round-bottom flask under stirring, 30.0 g (0.217
mol) 4-hydroxybenzoic acid were added to 135 mL of methanol,
followed by the dropwise addition of 45 mL of an aqueous
solution of potassium hydroxide (45 wt %). When the solution
became clear, 36.6 g of 5-bromo-1-pentene (0.254 mol) was
added dropwise to the solution using an addition funnel. After
refluxing for 16 h, the mixture was then cooled to room
temperature and poured into 500 mL of deionized water to
form a transparent yellow solution. An organic phase was then
extracted with 100 mL diethyl ether (alternatively, hexanes)
three times. To the aqueous phase, 100 mL of a 37% hydro-
chloric acid solution were added, leading to the precipitation
of the desired product. The precipitate, a white solid, was
collected after filtration and further recrystallized from etha-
nol. The purity and structure were confirmed by GC-MS and
1H NMR. The yield before recrystallization was calculated to
be about 50%. 1H NMR in D6-acetone gave δ: 10.9 (1H, d),
7.98 (2H, d), 7.02 (2H, d), 5.88 (1H, m), 5.00 (2H, m), 4.11 (2H,
t), 2.25 (2H, m), 1.90 ppm (2H, m).
Differential scanning calorimetry (DSC) experiments were
performed on a TA Instruments Q100 apparatus with heating
and cooling rates of 10 °C/min, unless otherwise stated, and

Macromolecules, Vol. 38, No. 10, 2005
Phase Transitions via Mixed-Mesogen LC Polymers 4105
Scheme 1. Schematic Representation of the Two Routes (I: Nucleophilic substitution with acid chloride, and II:
Esterification under DCC catalysis) Followed for the Preparation of the Two Diene Liquid Crystalline
Monomers: 1,4-bis[4-(4-Pentenyloxy)Benzoyl]-hydroquinone (M1), and
2-tert-butyl-1,4-bis[4-(4-pentenyloxy)Benzoyl]hydroquinone (M2)
Synthesis of 5-Pentenyloxybenzoic Acid Chloride (1b).
In a round-bottom flask, 15.0 g (72.7 mmol) of 5-pentenyloxy-
benzoic acid, 30 mL SOCl2, and several drops of DMF (as
catalyst) were added, and the solution was refluxed for 8 h.
Most of the SOCl2 was then removed by ambient pressure
distillation to yield a viscous, dark-yellow oil-like liquid. The
latter was further purified by vacuum distillation. The boiling
point of the final product, a pale-yellow liquid, was found to
be 130 °C (0.75 mm Hg). The yield after distillation was
until stoichiometric equivalence had been reached. The result-
ing mixture was stirred for 4 h at T ) 0 °C and, subsequently,
for 8 h at room temperature. The reaction solution remained
homogeneous (clear) throughout the reaction. After removal
of solvent by evaporation, the raw product (a waxlike brown
solid) was collected and purified by column chromatography
using silica gel as the stationary phase and a mixture of ethyl
acetate and hexanes (1:7 v/v) as eluent. 1H NMR in D6-acetone
gave δ: 8.16 (4H, m), 7.33 (1H, s), 7.22 (2H, t), 7.14 (4H, m),
5.95 (2H, m), 5.04 (4H, m), 4.16 (4H, t), 2.28 (4H, m), 1.93 (4H,
m), 1.39 (9H, s) ppm (see Figure 1).
Synthesis of M2 by route II. To a 250 mL Erlenmeyer
flask, 8.0 g (0.039 mol) of 1, 8.0 g (0.039 mol) of 1,3-dicyclohexyl
carbodiimide (DCC) (note: 1 and DCC have the same molec-
ular weight), 3.23 g of prepurified tert-butylhydroquinone
(0.0194 mol), 0.430 g of 4-(dimethylamino)pyridine (DMAP)
(3.52 mmol), and 80 mL of anhydrous dichloromethane were
added. The solution quickly became milky as the reactions
1
measured to be around 52%. H NMR in CDCl3 gave δ: 8.06
(2H, d), 6.95 (2H, d), 5.84 (1H, m), 5.04 (2H, m), 4.06 (2H, t),
2.25 (2H, m), 1.93 ppm (2H, m).
Synthesis of M2 by Route I. This synthetic route is similar
to that reported by Shiota and Ober,12 yet with different
compounds and purification procedures. First, 7.90 g (35.2 mol)
of 1b was added to a 250-mL round-bottom flask and cooled
to 0 °C with the aid of an ice bath. Then, a solution of tert-
butylhydroquinone in pyridine (0.75 M) was added dropwise
Figure 1. Schematic representation of the chemical structures of M1 (top) and M2 (bottom) along with the chemical shifts of
1
each hydrogen as measured by H NMR.

4106 Rousseau et al.
Macromolecules, Vol. 38, No. 10, 2005
Scheme 2. Schematic Representation of the Polymerization of Siloxane-based Liquid Crystalline Polymers and
Preparation of Their Blendsa
a The mesogenic units are: (M1) 1,4-bis[4-(4-pentenyloxy)benzoyl]hydroquinone and (M2) 2-tert-butyl-1,4-bis[4-(4-pentenyloxy-
)benzoyl]hydroquinone. The flexible spacers are: (S1) 1,4-bis(dimethylsilyl) benzene, and (S2) 1,1,3,3,5,5-hexamethyltrisiloxane.
The composition of each blend is indicated by the mole fraction of M1 and M2, x and y, respectively, as introduced by the constitutive
homopolymers.
Scheme 3. Schematic Representation of the Hydrosilation Reaction Followed for the Copolymerization of
Siloxane-based Liquid Crystalline Copolymers Incorporating M1 and M2, Respectively, as Mesogenic Units and a
Hydride-terminated Poly(dimethylsiloxane) (DP ) 8), S3 as Flexible Spacera
a The composition of the copolymers in terms of molar percentage of mesogenic units is indicated by x′ and y′.
proceeded and was continuously stirred for 48 h. At the end
of the reaction, a white solid as a byproduct was removed by
filtration, resulting into a clear brown solution of the desired
product in dichloromethane. After the removal of CH2Cl2 by
evaporation, the waxlike, brown raw product was collected.
The raw material was then purified by column chromatogra-
phy as described for route I.
ing the structure and good purity of the final polymers. For
comparison purposes, a cross-linked, mesogen-free analogue
of the copolymers was prepared by polymerizing S3 with higher
molecular weight R,ω-divinyl-terminated poly(dimethylsilox-
ane) and 12.5 mol % of CL. Below, we give a specific example
for the polymerization of P(50M1-co-50M2).
Synthesis of P(50M1-co-50M2). In a Schlenk flask, the
monomers were initially introduced, such as 0.0971 g of M1
(0.200 mmol) and 0.1093 g of M2 (0.202 mmol). Under nitrogen
flow, anhydrous dichloromethane was added to the mesogen
mixture to the extent of 2 mL/mmol of mesogens as well as
the catalyst (1 drop/0.4 mL). The mixture was heated to 45 to
50 °C, and the hydride-terminated poly(dimethylsiloxane) was
added dropwise until stoichiometric equivalence had been
safely exceeded (0.68 mL, 1.09 mmol). The reaction is allowed
to proceed for 24 h at 45-50 °C. The copolymer is purified by
precipitation in methanol and dried in a vacuum oven over-
night at 50 °C.
Polymerizations. The polymerization and copolymeriza-
tion routes are shown in Schemes 2 and 3, respectively. We
have adopted a well-established hydrosilation reaction8,11,13
catalyzed by platinum(0)-1,3-divinyl-1,1,3,3-tetramethyldi-
siloxane complex in xylenes. Unless stated otherwise, all the
polymerizations were conducted in anhydrous dichloromethane
(CH2Cl2) at 45-50 °C for 24 h in a Schlenk flask under stirring.
Both the catalyst and solvent were chosen in an attempt to
yield cleaner, more efficient reactions, i.e., higher yields and
higher resulting molecular weights.14 Attempted polymeriza-
tions using another catalyst, dichloro(cycloocta-1,5-diene)-
platinum in dichloromethane (Pt(II)), as well as other solvents,
including tetrahydrofuran and toluene, proved inferior. After
a solution of mesogen in anhydrous dichloromethane (1 mmol
mesogen:2 mL solvent) and Pt(0) catalyst was brought to 50
°C, the disilane was added slowly until equimolar stoichiom-
etry was achieved. The reaction was allowed to proceed for 24
h, following which the solutions were diluted about 3 times in
CH2Cl2 and further precipitated in hexanes or methanol. The
polymers (PMiSj, where i ) 1, 2 and j ) 1, 2, 3; see materials
and methods, above) and copolymers (P(x′M1-co-y′M2)) were
dried overnight at 40 °C under vacuum before further char-
acterization. 1H NMR of the homopolymers were run, confirm-
Preparation of Blends. Binary blends of PMiSj homopoly-
mers were prepared by codissolution in a mutual solvent and
subsequent drying. Specifically, solutions in dichloromethane
of PM1S1 (0.135 M, using mol/repeat unit) with PM2S1 (0.135
M) or PM1S2 (0.086 M) with PM2S2 (0.042 M) were combined
in varying proportions to yield blend compositions spanning
0-100 mol % (moles of repeat unit) in 10% increments. Note
that the mol % values for this blend are quite similar to wt %.
After dissolving the polymers to form clear and homogeneous
solutions, the solvent was evaporated, first at ambient condi-
tions and subsequently under vacuum at T ) 40 °C for 12 h.

Macromolecules, Vol. 38, No. 10, 2005
Phase Transitions via Mixed-Mesogen LC Polymers 4107
Figure 2. DSC traces at 10 °C/min of: (a) 1,4-bis[4-(4-pentenyloxy)benzoyl]hydroquinone (M1), and (b) 2-tert-butyl-1,4-bis[4-(4-
pentenyloxy)benzoyl]hydroquinone (M2). For each trace shown, thermal history has been erased through an initial melting at
250 °C for M1 and 150 °C for M2 (except for the sample crystallized from solution) followed by a cooling at 10 °C/min down to 0
°C. (a) On heating, M1 crystals melt into a nematic phase at 136.6 °C (∆H ) 78.24 J/g), which clears at 229.5 °C (∆H ) 5.50 J/g).
(b) On heating, M2 crystals formed from solution melt into an isotropic phase at 105.9 °C (∆H ) 73.08 J/g), whereas M2 crystals
formed from the melt, after annealing at 0 °C for 3 h (indicated times), melt into a nematic phase at 80.6 °C (∆H ) 50.97 J/g),
which further clears at 91.4 °C (∆H ) 3.55 J/g). The cooling trace is identical for all runs and only displays the isotropic-to-
nematic transition.
Figure 3. POM images under crossed polarizers of the mesophases exhibited by M1 and M2 under varying temperature. The
letters refer to the mesophases exhibited by M1 and M2 as depicted in Figure 2. M2 not annealed on heating at (a) 56.8, (b) 83.4,
(c) 86.0, and (d) 91.2 °C. (e) M2 crystals formed from the melt after annealing for 3 h at room temperature. M1 on cooling from (t)
the isotropic phase at 240 °C to (u) the nematic phase at 237.1 °C to (v) the crystalline phase at 71 °C. The 200 µm scale bar
refers to every image except (e).
Results and Discussion
105.9 °C (crystal-isotropic), while, in the latter case,
two transitions are evidenced by two endotherms oc-
curring at about 80.6 (crystal-nematic) and 91.4 °C
(nematic-isotropic), the crystal-nematic transition tak-
ing time at room temperature to develop (Figure 2b, 180
min). A combination of DSC analysis with polarizing
optical microscopy (POM) (see Figure 3) and wide-angle
X-ray diffraction (WAXD) studies (not shown here)
permitted the determination of the nature of these
transitions as indicated. In particular, Figure 3 shows
selected POM images acquired at temperatures indi-
cated by letters that correspond to particular regions
on DSC traces of Figure 2. POM images “a” through “d”
Mesogen Characterization. The phase behavior of
both M1 and M2 were probed using differential scanning
calorimetry (DSC), the results of which are presented
in Figure 2. Upon heating at 10 °C/min, M1 displays
(Figure 2a) transitions in the form of sharp endothermic
peaks at 136.6 (crystal-nematic) and 229.5 °C (nemat-
ic-isotropic), the first featuring a latent heat ap-
proximately 10-fold the second. By comparison, M2
appears to have a more complex behavior (Figure 2b),
with differences depending on sample history; precipi-
tated from solution or cooled from the molten state. In
the former case, M2 exhibits a single endotherm at

4108 Rousseau et al.
Macromolecules, Vol. 38, No. 10, 2005
Table 1. Summary of the Homopolymers’ Physical,
Thermal, and Phase characteristics
reveal a nematic threaded texture (a, b) for M2 that
clears through a narrow nematic-isotropic biphase (c)
to an isotropic liquid (d). If crystallized, however, M2
exhibits featherlike crystals that grow radially with
significant branching (e). Cooling M1 from the isotropic
phase (t) leads to a nematic Schlieren texture (u),
populated primarily by wedge-type disclination lines,
and subsequently to a dark (scattering) crystalline
phase (v).
WAXD patterns for both M1 and M2 were consistent
with POM and DSC findings, while allowing clear
identification of low-temperature phases. In particular,
M1 exhibited a multitude of crystalline reflections at T
) 30 °C (supplementary data), consistent only with a
crystalline phase.15-16 Heating to T ) 145 °C, a condi-
tion between the endotherms of Figure 1a and for which
a nematic Schlieren texture is evident in Figure 2, only
an amorphous halo in WAXD was observed (d-spacing
) 4.86 Å), allowing conclusion of a nematic phase with
local order (birefringent domains with domain size 1 <
a < 50 µm) but macroscopic disorder (WAXD halo). Like
M1, M2 is crystalline at room temperature, as evidenced
by numerous diffraction peaks in WAXD observations.
Interestingly, the crystal structure of M2 depends on
preparation conditions, with clear differences in peak
positions and d-spacings for solution-crystallized mate-
rial and melt-crystallized material.
samples
Mh w (kDa)
PDI
transition temperatures (°C)
PM1S1
PM2S1
PM1S2
PM2S2
22.2
60.9
1.73
1.59
1.91
2.25
SmC + Cr 152.2, N 158.1 I
Ig 57.4 I
66.5
Sm 135.5, SmC 169.8 I
163.2
LCg 18.5, LCa 38.9 I
a After annealing for 8 h at 10 °C. Phases are indicated as
follows: Cr: crystalline phase. SmC: smectic-C. Sm: undefined
higher-order smectic phase. N: nematic. LC: undefined liquid
crystalline phase. I: isotropic. LCg, Ig: liquid crystalline or
isotropic glass. Polymers bearing spacers Si and mesogens Mi are
defined as PMiSi, where M1 is 1,4-bis[4-(4-pentenyloxy)benzoyl]
hydroquinone, M2 2-tert-butyl-1,4-bis[4-(4-pentenyloxy)benzoyl]
hydroquinone, S1 1,4-bis(dimethylsilyl) benzene, and S2 1,1,3,3,5,5-
hexamethyltrisiloxane.
Polymers and Blends Characterization. Scheme
2 gives a representation of the synthetic route for the
homopolymerization of M1 and M2 with two different
silicon-based flexible spacers, namely, 1,4-bis(dimeth-
ylsilyl)benzene (S1) and 1,1,3,3,5,5-dimethyltrisiloxane
(S2), along with the blending procedure. Although the
synthesis of M1 has been described by other groups,
preparation of the tert-butyl-substituted counterpart has
not been reported before, and only one report on its
incorporation in polymers has appeared from our group.17
Here, we show the results obtained upon polymeriza-
tion, via conventional hydrosilation reaction, of this new
mesogenic unit with different siloxane and silane spac-
ers, PM2S1 and PM2S2, as well as their behavior when
blended with their M1 polymeric homologues, PM1S1
and PM1S2. The physical characteristics of the ho-
mopolymers, on the basis of DSC, POM, and WAXD
analyses, are summarized in Table 1. In all cases,
relatively high molecular weights were achieved, with
Mh w ranging from 22 to 163 kDa, and yields spanning
from 45 to 92%. As expected from previous studies,13,18-20
the glass transition, Tg, of the M2-based homopolymers
decreased considerably upon polymerization with longer
and/or more flexible spacers ranging from about 56 °C,
when polymerized with 1,4-bis(dimethylsilyl)benzene,
to 21 °C when polymerized with 1,1,3,3,5,5-dimethyl-
trisiloxane, to Tg ) -17 °C when polymerized with the
longer hydride-terminated poly(dimethylsiloxane) spacer
(S3) (described below).
DSC thermograms for PM1S1 and PM2S1 are shown
as the top and bottom traces of Figure 4a, respectively,
while those for PM1S2 and PM2S2 are shown as the top
and bottom traces of Figure 4b. Focusing first on the
lower temperature ranges of each DSC trace, one can
see the dramatic reduction in glass transition temper-
ature observed for the two M2-based homopolymers
(bottom traces) on increasing the spacer flexibility from
that of S1 to S2. On the basis of DSC analysis (Figure
4), POM, and WAXD (data not shown), we found that
polymerizing with S1 gave rise to both a smectic and a
nematic phase, while using S2 as a flexible spacer
prevented the formation of a nematic phase and led to
materials solely exhibiting smectic phases of different
degrees of order. In particular, the top trace of Figure
4a, that of PM1S1, shows an initial endothermic transi-
tion (Tm ) 124.6 °C) that grows upon annealing near
but below this temperature, consistent with a melting
transition. This melting transition is followed by a sharp
transition from a smectic-C to a nematic phase and
subsequent isotropization with a nematic window 28 °C
in breadth. In contrast, the top trace of Figure 4b,
PM1S2, features two sharp endothermic transitions, but
Thus, we have observed that, at low temperatures,
M1 exists in a crystalline form, which melts to a nematic
phase at 136.6 °C and further clears to an isotropic
phase at 229.5 °C. These observations stand in contrast
with the conclusions made by Kossmehl et al. regarding
their characterization of the same compound.10 In their
report, the authors described a smectic-to-nematic
transition followed by a nematic-to-isotropic transition
at similar transition temperatures, largely on the basis
of DSC analysis. Although we are unable to explain the
difference in our results, the ability of our material to
crystallize to such a large extent (Figure 2a) suggests
very high purity that may not have been achieved in
the other study. In the case of M2, we observed that
crystallization is more complex and seems to yield two
different crystalline forms whether nucleated in solution
or upon cooling from the melt. While crystals formed
from solution directly melt to an isotropic phase at a
comparatively high Tm () Tki) ) 105.9 °C, crystals that
form on cooling subsequently melt to a nematic phase
at a significantly lower Tkn ) 80.6 °C, i.e., one-time
monotropism, and clear to an isotropic phase at Tni
)
91.4 °C. Moreover, crystallization of M2 following a first
melting is greatly retarded, such that cooling to T ) 0
°C at 10 °C/min does not result in any crystallinity
(Figure 2b, cooling). Instead, annealing at 0 °C for 30
min or longer is required to complete crystallization.
The similar chemical structures of M1 and M2,
combined with their contrasting phase behavior, sug-
gested the promise of blending their polymeric deriva-
tives to allow preparation of new thermotropic liquid
crystalline materials with both low and tunable transi-
tion temperatures as dictated by prescribed composition.
Thus, in the following sections, we present the synthesis
of main-chain siloxane-based liquid crystalline poly-
mers, blends, and copolymers, each incorporating vary-
ing amounts of M1 and M2 and using different silicon-
based flexible spacers. We further discuss their phase
behavior and structures in light of their potential use
in network form for thermally stimulated actuation.

Macromolecules, Vol. 38, No. 10, 2005
Phase Transitions via Mixed-Mesogen LC Polymers 4109
Figure 4. DSC traces of (a) various xPM1S1/yPM2S1 blends, and (b) various xPM1S2/yPM2S2, on second heating at 10 °C/min.
From bottom to top, the curves indicate increasing content of the M1-based homopolymer in the blends via increments of 20 mol
%, so that (i), (j), (t), and (u) correspond to pure PM2S1, PM1S1, PM2S2, and PM1S2, respectively. The trace (t) of PM2S2 corresponds
to an annealed sample at room temperature so as to clearly demonstrate the existence of ∆HLC-I. (Sm: undefined smectic phase
of high order, smectic phase. SmC: smectic-C. N: nematic. I: isotropic; LCg. Ig: liquid crystalline glass or isotropic glass. Cr:
crystalline phase).
this time more widely separated and featuring a low-
temperature smectic f smectic-C f isotropic phase
sequence confirmed with hot-stage WAXD and POM. An
attempt to describe the low-temperature smectic phase
is given below. Hence, the S2 spacer has a greater
propensity toward smectic phase formation than the S1
spacer.
layer, regardless of inclination, contains mesogens tilted
at an angle of 40.3° with respect to the smectic normal,
using a fully extended repeat unit length computed to
be 35.6 Å.
Inspection of the lower DSC traces of Figure 4a and
b reveals that both homopolymers PM2S1 and PM2S2
(those solely incorporating M2 as mesogen) exhibit a
single thermal transition being a well-defined glass
transition temperature at 57.4 and 19.1 °C, respectively.
Hot-stage POM of the same samples at temperatures
above these Tg values (data not shown) indicated
complete absence of birefringence, confirming the iso-
tropic nature of the materials above Tg. However, on
annealing below Tg, PM2S1 exhibits an additional
endotherm slightly higher than Tg indicative of a liquid
crystalline-to-isotropic transition. Apparently, in ho-
mopolymers formed with these particular spacers, S1
and S2, the pendant t-butyl group is bulky enough to
suppress or delay alignment in a nematic phase or
ordering into smectic layers, respectively. This behavior
is in stark contrast with thermotropic LCPs with the
same mesogen, but bearing substantially shorter, all-
hydrocarbon spacers of eight (both saturated22 and
unsaturated23) and ten24 methylene units. For these
polymers with shorter spacers, crystallization is still
suppressed, but a nematic phase is observed with
nematic-isotropic transition much higher than those
of PM2S1 and PM2S2, being in all cases higher than 170
°C. Thus, we conclude that the spacer itself plays a
leading role in determining phase ordering, at least
when paired with a mesogen featuring a bulky pendant
group like the t-butyl group used in this study.
While we suspect the nature of the low-temperature
smectic phase of PM1S2 to be of crystalline order
because of the sharp reflection at 4.14 Å as well as the
shallower reflections at lower angles (6.30, 6.70, and
7.56 Å) obtained from X-ray measurements (Supporting
Information), we were not able to specify its structure
with certainty. We speculate that it is a G-like phase21
(a type of smectic), whose POM texture and X-ray
diffraction pattern are consistent with the ones obtained
in this study; however, we presently classify it as
“smectic (Sm)” until its nature can be defined with more
confidence. The d-spacing values characteristic of the
smectic layering of PM1S1 and PM1S2 were determined
to be 24.3 and 26.8 Å, respectively (Supporting Informa-
tion), with the associated reflection being nonisotropic
with multiple off-meridian peaks for fiber samples. In
the case of PM1S1, the WAXD fiber pattern is consistent
with a chevron microstructure wherein mesogens are
oriented in the fiber direction and further assembled
into inclined smectic layers. The smectic d-spacing value
(24.3 Å) is in good agreement with the calculated value
of 26.2 Å obtained using the extended conformation of
the repeat unit constituting each smectic layer (33.6 Å)
tilted at an angle of 38.8° to the smectic normal, in
accordance with the measured angle between this
normal and the stretching direction (director).
Upon blending PM1S1 with PM2S1 (1,4-bis(dimeth-
ylsilyl)benzene spacer) and PM1S2 with PM2S2 (tri-
siloxane spacer), we expected to observe miscibility
given their similarity in chemical structure, but instead
we found the pairs to be immiscible. Focusing on the
intermediate composition traces in Figure 4a and b
reveals simple additivity of the homopolymer traces (top
The microstructure of PM1S2, while similar to PM1S1,
is more complex, revealing 10 low-angle (smectic-like)
reflections (Supporting Information), suggesting a com-
plex chevron pattern that involves multiple inclination
angles (0, 17.6, and 49.7°) of the smectic layers, but with
constant layer spacing of 26.9 Å. Thus, each smectic

4110 Rousseau et al.
Macromolecules, Vol. 38, No. 10, 2005
Figure 5. Close inspection of the last two endothermic
transitions (TSN ≈ 152 °C and TNI ≈ 159 °C) of several PM1S1/
PM2S1 blends: (a) 90/10, (b) 80/20, (c) 70/30, (d) 60/40, (e) 50/
50, (f) 40/60. Inset: the relative variation of each endothermic
enthalpy as a function of the blend composition. This shows
the propensity of PM1S1-rich blends to form a smectic phase,
in contrast to PM2S1-rich blends, while all compositions in this
range maintain a similar nematic phase.
Figure 6. Close inspection of the last two endothermic
transitions (TSS ≈ 135 °C and TSI ≈ 169 °C) of several PM1S2/
PM2S2 blends: (a) 90/10, (b)80/20, (c) 70/30, (d) 60/40, (e) 50/
50, (f) 40/60, (g) 30/70, (h) 20/80. Inset: the relative variation
of each endothermic enthalpy as a function of the blend
composition. Because of the similar nature of the two smectic
phases exhibited by these blends, their change in composition
affects only very slightly the persistence of each phase relative
to the other.
and bottom traces) with weighting of the feature sizes
in proportion to the amount of that component. For
example, the 60PM1S1/40PM2S1 blend shows a Tg
essentially unmodified from the PM2S1 homopolymer
(54.5 and 57.4 °C, respectively), while the combined
endothermic latent heats for the smectic-nematic and
nematic-isotropic transitions are 55.5% of pure PM1S1.
Similar findings are observed for other blends in this
dictating the mixing behavior of these LCPs. This is in
accordance with the general observation for small-
molecule liquid crystals that alike phases broadly mix
(often used to identify the phase of a new LC) indepen-
dent of composition.
Although these new blends exhibit relatively low
transition temperatures, including low glass transitions,
they suffer from an inability to tailor transition tem-
peratures with composition because of immiscibility. As
our application goal is to incorporate new LCPs into
networks for soft actuation with variable transition
(triggering) temperature, another approach was needed.
Restricting ourselves to incorporation of M1 and M2
mesogens, because of their desirable breadth in phase
behavior, and the use of silicon-type spacers that enable
low glass transition, we chose to prepare and study the
behavior of analogous copolymer-based systems. Both
the synthesis and properties of these new materials are
therefore described in the following section.
Copolymer Characterization. As mentioned above,
copolymerizing instead of blending as an approach
toward mixing property contributions of mesogens M1
and M2 was expected to yield liquid crystalline materials
of greater interest with respect to their end-use applica-
tions. In particular, this route was expected to yield
homogeneous systems with transition temperatures
that would vary continuously with composition. Ad-
ditionally, this one-step routine (in contrast to blending)
is made inherently simpler by avoiding the solvent
blending step. The selection for the silicon-based spacer
to be incorporated into the preparation of M1/M2 co-
polymers was based upon our observations for PM2Sn
homopolymers: increasing the flexibility from that of
1,4-bis(dimethylsylil)benzene to 1,1,3,3,5,5-hexameth-
yltrisiloxane decreased Tg fom 57 to 19 °C. Therefore,
we postulated that increasing further the flexibility
(length) of a siloxane-based spacer, to include eight
series with the ratio ∆Hblend/∆HPM S , exhibiting direct
1
proportionality to the weight percen1tage PM1S1 in the
blend. On closer inspection, we observe in Figure 5 that
the nematic-isotropic transition enthalpy drops less
quickly with increasing PM2S1 in the blend than does
the smectic-nematic enthalpy, indicating that the
amorphous component is more disruptive to smectic
ordering than to nematic alignment for the case of the
S1 spacer.
Similar behavior prevails in the case of PM1S2/PM2S2
blends, DSC results of which are shown as the inter-
mediate traces of Figure 4b. In these cases, the lower
PM2S2-derived Tg remains virtually unchanged at 19
°C for all of the blends, and the two endothermic
transitions (undetermined smectic f smectic-C and
smectic-C f isotropic) decrease in enthalpy in nearly
direct proportion to the amount of amorphous PM2S2
added (Figure 6). However, in contrast to the PM1S1/
PM2S1 blends, we found that the enthalpy of the higher
temperature transition (smectic-C f isotropic) decreases
with increasing PM2S2 content with a similar sensitivity
to that of the lower-temperature smectic f smectic-C
transition (Figure 6). We reason that because of both
the similarity of the liquid crystalline phases and the
fact that PM2S2 itself exhibits liquid crystallinity (in
contrast to PM2S1), the change in blend composition
does not affect the persistence of each smectic phase
relative to the other. In light of these findings, it seems
that the difference in phase behavior between M1- and
M2-bearing LCPs dominates over chemical similarity in

Macromolecules, Vol. 38, No. 10, 2005
Phase Transitions via Mixed-Mesogen LC Polymers 4111
Table 2. Summary of Copolymer Properties
sample name
Mh w (kDa) PDI “Tg like” (°C) TSI (°C)a
P(M1S3)
63.5
56.7
47.9
43.7
44.5
58.3
42.8
59.0
2.16
1.46
1.53
1.48
1.55
1.53
1.51
1.58
49.3
47.2
41.8
30.7
17.7
7.6
93.0
72.9
68.4
60.8
53.9
49.2
18.3
-0.1
P(90M1-co-10M2)
P(80M1-co-20M2)
P(70M1-co-30M2)
P(60M1-co-40M2)
P(50M1-co-50M2)
P(25M1-co-75M2)
P(M2S3)
-3.1
-16.7
a TSI: smectic-to-isotropic transition temperature.
siloxane units, would lower Tg enough to yield properties
at near room temperature that were independent of
proximity to Tg. Thus, we selected a hydride-terminated
poly(dimethylsiloxane) (DP ) 8) (S3, Scheme 3) as the
spacer used in the preparation of M1/M2 copolymers.
The physical, optical, and thermal properties of these
new copolymers are discussed below. In addition, their
properties relevant to eventual use in networks as soft
actuators will be briefly introduced.
Copolymers were successfully synthesized using hy-
drosilation (see Scheme 3), yielding molecular weights,
Mh w, ranging from 42 800 to 63 500 g/mol. The copoly-
mers are designated as P(x′M1-co-y′M2). where x′ and
y′ indicate the molar percentages incorporated in poly-
merizations with a stoichiometric equivalence of S3
implied. The molecular characteristics, physical proper-
ties, and phase behavior of these new materials were
studied using GPC, hot-stage POM, and DSC, the
results of which are summarized in Table 2 and Figure
7. Immediately, we noticed that incorporation of S3 as
a spacer shifts the glass transition to subzero temper-
atures as low as T ) -17 °C. In Figure 7, the heating
DSC trace of the mesogen-free analogue (similar to
cross-linked PDMS) is also given. A clear melting
transition near -50 °C is observed but no glass transi-
tion for T > -80 °C. Apparently, the PDMS spacer of
the copolymers is too short to permit the crystallization
evident in high-molecular-weight PDMS. Moreover, it
appears as if the glass transition exhibited by the
copolymers arises from the mesogens themselves and
not from their “soft” siloxane-based component. At
higher temperatures, one can see for the copolymers
that a broad endothermic transition shifts to lower
temperatures continuously with increasing M2 incor-
poration. This continuous adjustment in transition
temperature is a clear indication of molecular-level
homogeneity of mesogen distribution in the materials,
in contrast to the significant heterogeneity indicated for
the analogous blends.
The phase transition behavior of this series of copoly-
mers (Figure 7) is complex, consisting of a broad (20 <
∆T < 50 °C) and complex endothermic transition but
exhibiting a smectic-C phase at low temperatures, as
determined by X-ray studies on oriented samples, and
an isotropic phase at high temperatures. Copolymer
phase transition behavior is further complicated by the
presence of the glass transition that is superimposed
on the clearing transition, as characterized by the
significant shift in baseline that occurs near or within
the transition. In addition to thermal behavior, Figure
7 also displays hot stage POM images obtained for a
specific copolymer, P(70M1-co-30M2), observed under
crossed polarizers. This particular copolymer features
a relatively simple (though broad) clearing transition,
again superimposed on the glass transition that is seen
in POM as a transition from a granular birefringent
Figure 7. Thermal behavior of the siloxane-based liquid
crystalline copolymers, P(x′M1-co-y′M2), as monitored by DSC
measurements on heating at 5 °C/min as a function of M1 and
M2. For all samples, the thermal history has been erased
through an initial melting at 200 °C followed by a cooling ramp
at 5 °C/min. The POM images correspond to the temperature
region demarked by arrows for sample P(70M1-co-30M2). The
top trace corresponds to a mesogen-free analogue, revealing
the absence of any glass transition in the temperature window
probed for the study of the copolymers.
texture to a dark, featureless micrograph indicating
optical isotropy. Physically, the material is gel-like at
room temperature and liquid-like above TsmC-I, in this
case, 72 °C. In a recent report9 concerning P(70M1-co-
30M2) cross-linked with a 10-mol % tetrafunctional
cross-linker (tetrakis(vinyldimethylsiloxy) silane), Rous-
seau et al. found that this behavior changes to a
stepwise decrease in shear rubber modulus from about
10 MPa at room temperature to 100 kPa above T ) 72
°C.
As mentioned above, the low-temperature phase
exhibited by copolymers of M1 and M2 with the S3 spacer
is smectic-C. Such phase order was established by
gathering transmission WAXD data on samples pro-
cessed into fibers drawn from the melt and cooled to
room temperature. Figure 8 shows the two-dimensional
(2D) WAXD pattern for P(70M1-co-30M2) prepared in
this way. In this case, we observed a pattern charac-
teristic of a smectic-C phase, as indicated by the clear
splitting of the low-angle smectic peaks into two pairs
of peaks (four peaks total) separated azimuthally by
roughly 45°. The second-order reflections of these smec-
tic peaks are clearly visible (d-spacing ) 16.1 Å) and
indicate a significant long-range order of the tilted
smectic layers (d-spacing ) 31.9 Å). At wider angles (d-
spacing ) 5.7 Å), strong and moderately sharp equato-

4112 Rousseau et al.
Macromolecules, Vol. 38, No. 10, 2005
line materials toward spontaneously reversible actua-
tion, preliminary studies in our group have shown that
new main-chain liquid crystalline elastomers based on
a similar copolymer chemistry displayed promising
results in the area of shape memory,9 i.e., one-way
actuation, and were the subject of additional studies,
more to be reported. In addition, we showed that these
new LC copolymers exhibit a glass transition that is
superimposed on their complex clearing transition. We
have reported earlier9 that this combination allowed for
the fixing of large strains and recovery of an equilibrium
shape when integrated into a shape memory design.
Figure 8. Two-dimensional WAXD pattern at room temper-
ature of a P(70M1-co-30M2) fiber drawn from the melt (draw-
ing direction vertical). The right-hand side of the figures shows
the interpreted microstructure of the underlying smectic-C
phase indicated by the X-ray diffraction pattern. The two
mesogens M1 and M2 are represented as dark and white rods,
while the main-chain siloxane backbone is omitted for clarity.
Conclusions
In this report, we first show the synthesis and
characterization of both a new mesogen, namely, 2-tert-
butyl-1,4-bis[4-(4-pentenyloxy)benzoyl]hydroquinone, M2,
with low transition temperatures (Tm ) 80 °C, TNI
)
91 °C) and its already known unsubstituted homologue,
namely, 1,4-bis[4-(4-pentenyloxy)benzoyl]hydroquinone,
M1, exhibiting much higher transition temperatures (Tm
) 136 °C, TNI ) 229 °C). We believe that such difference
in phase transition behavior is exclusively due to the
tert-butyl substitution and its resulting steric hindrance
on packing ability. Despite their very similar chemical
structures, the drastically different phase transitions
of M1 and M2 suggested that mixing would allow
targeting of specific properties, specifically transition
temperatures. Therefore, we showed the successful
synthesis and preparation of new main-chain siloxane-
based liquid crystalline polymers and blends incorporat-
ing M1 and M2 and using varying silicon-based flexible
spacers. The blends obtained possessed either smectic-
isotropic or smectic-nematic-isotropic mesophase se-
quences depending on the nature of the flexible spacer.
Even though relatively low Tg and TNI were achieved
by blending, 20 °C and 175 °C, respectively, the im-
miscibility exhibited by the blends redirected us toward
a copolymerization route using similar chemistry. Thus,
copolymers were synthesized using hydrosilation to high
molecular weights, Mh w, ranging from 29 to 87 kDa, and
liquid crystallinity over wide temperature windows as
governed by their chemical composition. The transition
temperatures of the copolymers were tuned via compo-
sition changes such that smectic-isotropic clearing
ranged from 0 to 90 °C and is more or less superimposed
on the glass transition that varies from about -17 to
51 °C. Such tailorability of low glass transition and
isotropization temperatures renders these new materi-
als adequate candidates for further incorporation into
liquid crystalline elastomers for soft actuation devices.
rial reflections indicate strong, yet noncrystalline, in-
termesogen correlations within the smectic layers. At
intermediate angles, an amorphous ring is superim-
posed with meridional layer lines corresponding to the
siloxane spacers (intermediate d-spacings, 7.5 and 6.8
Å).
On the basis of these features in 2D WAXD, we
envision these copolymers to exist at low temperatures
in a chevronlike smectic-C layering, shown schemati-
cally in Figure 8. The measured layer spacing of 31.9 Å
agrees well with the theoretical value of 34.0 Å, com-
puted on the basis of a fully extended chain conforma-
tion arranged into tilted smectic layers adopting the
measured inclination of 45° with respect to the stretch-
ing direction (see Supporting Information). The tilt
angle measured for all copolymers examined in this way
reside around a value of 44.9 ( 0.7°, which is quite
reasonable for such a smectic-C arrangement. Thus, the
P(x′M1-co-y′M2) copolymers have been shown to dem-
onstrate broad ranges of liquid crystallinity, in some
cases more than 100 °C, but also to feature relatively
low isotropization temperatures that can be tuned via
chemical composition to vary between 0 and 90 °C.
Such observed desirable characteristics for mixed-
mesogen LCPs with siloxane-bearing spacers (P(x′M1-
co-y′M2)) have been the goal of recent studies incorpo-
rating LC materials for actuation purposes,5,7,9,25 where
low modulus and low triggering temperatures are
required to enable certain biomedical devices.26-28 It is
important to note that actuation phenomena in LCEs
are governed mainly by: (i) their isotropization/anisotro-
pization capabilities, which in turn dictate the condi-
tions for actuation to occur and to what extent, and (ii)
the transition temperatures, which trigger the mechan-
ical transitions of the resulting actuators. It is the
combination of these two effects that dictates the overall
performances of the system. As mentioned earlier,
clearing transitions for the present polymers (see Figure
7) span a wide range of temperatures. Thus, significant
potential exists for these new materials to be used (in
network form) as high performance soft actuators as
they offer a wide range of transition temperatures easily
tunable by varying the mesogen and flexible spacer
composition and, therefore, a variety of end-use applica-
tions.
Acknowledgment. P.T.M. acknowledges the sup-
port of an NSF CAREER Award, Grant CTS-00093880.
Supporting Information Available: One-dimensional
(1D) X-ray diffraction patterns for M2; the 1D and 2D X-ray
patterns along with the azimuthal integration at low 2θ angles
of both M1-based homopolymer fibers; a WAXD 2θ profile of a
copolymer fiber drawn from the melt; azimuthal profile of
WAXD low angle reflections, characteristic of the smectic-C
phase, and of the large angle reflections indicative of the
mesogenic ordering of the same copolymer fiber; 2D WAXD
patterns for a range of M1/M2 copolymers. This material is
available free of charge via the Internet at http://pubs.acs.org.
References and Notes
In addition to the experimental work done by re-
searchers such as Finkelmann, Keller, and Ratna, which
showed the immense potential of similar liquid crystal-
(1) Pugh, C.; Kiste, A. L. Prog. Polym. Sci. 1997, 22, 601-691.
(2) Hsu, C.-S. Prog. Polym. Sci. 1997, 22, 829-871.

Macromolecules, Vol. 38, No. 10, 2005
Phase Transitions via Mixed-Mesogen LC Polymers 4113
(3) Warner, M.; Terentjev, E. M. Liquid Crystal Elastomers;
Oxford University Press: London, 2003.
(17) Qin, H.; Chakulski, B. J.; Mather, P. T.; Constable, G. S.;
Coughlin, E. B. Polym. Prepr. 2003, 44, 40-41.
(4) Finkelmann, H.; Kock, H. J.; Rehage, G. Makromol. Chem.,
Rapid Commun. 1981, 2, 317-322.
(18) Aguilera, C.; Bartulin, J.; Hisgen, B.; Ringsdorf, H. Makromol.
Chem. 1983, 184, 253-262.
(5) Thomsen, D. L., III; Keller, P.; Naciri, J.; Pink, R.; Jeon, H.;
Shenoy, D.; Ratna, B. R. Macromolecules 2001, 34, 5868-
5875.
(19) Sirigu, A. Liq. Cryst. Polym.: Princ. Fundam. Prop. 1991,
261-313.
(20) Diaz, F.; Tagle, L. H.; Valdebenito, N.; Aguilera, C. Polymer
(6) Terentjev, E. M. J. Phys.: Condens. Matter 1999, 11, R239-
R257.
1993, 34, 418-422.
(21) Gray, G. W.; Goodby, J. W. Smectic Liquid Crystals: Textures
and Structures; Heyden: Chichester, UK, 1984.
(22) Lee, J.-H. Development of an Electrospinning Device for
Thermotropic Liquid Crystalline Polymer (TLCP) Melts,
Thesis, University of Connecticut, 2004.
(23) Qin, Q.; Chakulski, B. J.; Rousseau, I. A.; Chen, J.; Xie, X.-
Q.; Mather, P. T.; Constable, G. S.; Coughlin, E. B. Macro-
molecules 2004, 37, 5239.
(24) Mather, P.; Grizzuti, N.; Heffner, G.; Ricker, M.; Rochefort,
W. E.; Seitz, M.; Schmidt, H.-W.; Pearson, D. S. Liq. Cryst.
1994, 17, 811-826.
(7) de Gennes, P.-G. C. R.′ Acad. Sci., Ser. IIb: Mec., Phys.,
Astron. 1997, 324, 343-348.
(8) Donnio, B.; Wermter, H.; Finkelmann, H. Macromolecules
2000, 33, 7724-7729.
(9) Rousseau, I. A.; Mather, P. T. J. Am. Chem. Soc. 2003, 125,
15300-15301.
(10) Kossmehl, G.; Gerecke, B.; Harmsen, N.; Schroeder, F.; Vieth,
H. M. Mol. Cryst. Liq. Cryst. Sci. Technol., Sect. A 1995, 269,
39-53.
(11) Kossmehl, G.; Gerecke, B.; Harmsen, N.; Vieth, H. M.; Wolff,
D. Mol. Crys. Liq. Cryst. Sci. Technol., Sect. A 1998, 317,
1-21.
(25) Wermter, H.; Finkelmann, H. Liquid Crystalline Elastomers
(12) Shiota, A.; Ober, C. K. J. Polym. Sci., Part A: Polym. Chem.
1996, 34, 1291-1303.
as Artificial Muscles. e-Polymers [Online] 2001, 13.
(26) Irie, M. Shape Mem. Mater. 1998, 203-219.
(13) Tronc, F.; Lestel, L.; Boileau, S. Polymer 2000, 41, 5039-
5046.
(27) Liu, C.; Chun, S. B.; Mather, P. T.; Zheng, L.; Haley, E. H.;
Coughlin, E. B. Macromolecules 2002, 35, 9868-9874.
(14) Sellinger, A.; Laine, R. M.; Chu, V.; Viney, C. J. Polym. Sci.,
Part A: Polym. Chem. 1994, 32, 3069-3089.
(15) A complete crystal structure analysis is beyond the scope of
this work and is the focus of ongoing research.
(16) Broer, D. J.; Hikmet, R. A. M.; Challa, G. Makromol. Chem.
1989, 190, 3201-3215.
(28) Lendlein, A.; Kelch, S. Angew. Chem., Int. Ed. 2002, 41,
2034-2057.
MA048327Y