Effect of molecular parameters on thermomechanical behavior of side-on nematic liquid crystal elastomers

Article abstract of DOI:10.1016/j.polymer.2013.07.057

We investigated the structure-property relationship of liquid crystal elastomers (LCEs) obtained from a series of nematic side-on monomers. A new synthetic strategy was developed to obtain the acrylate monomers (n-ADBB), which gave us the opportunity to easily modify the spacer lengths of the monomers. Through magnetic field alignment, well-defined LCE micropillars were fabricated from the monomers by a method combining soft lithography and photopolymerization/photocrosslinking of the monomers and a crosslinker. The influence of structural parameters on the thermomechanical deformation of the microstructures was studied through microscopic observations. The study quantitatively revealed the correlation of thermomechanical behavior of the microstructured LCEs with the crosslinking density and length of the flexible spacer linking the mesogenic core to the backbone. With a proper control of the structural parameters, optimized performances such as large reversible contraction, good elasticity and mechanical robustness were demonstrated for this type of LCEs.

Full text of DOI:10.1016/j.polymer.2013.07.057

Polymer xxx (2013) 1e9
Contents lists available at ScienceDirect
Polymer
journal homepage: www.elsevier.com/locate/polymer
Effect of molecular parameters on thermomechanical behavior
of side-on nematic liquid crystal elastomers
,
b,
Renbo Wei a, Lingyun Zhou a, Yaning He a, Xiaogong Wang a , Patrick Keller
*
**
a Department of Chemical Engineering, Key Laboratory of Advanced Materials (MOE), Tsinghua University, Beijing 100084, People’s Republic of China
b Institut Curie, Centre de Recherche, CNRS UMR 168, Université Pierre et Marie Curie, 26 rue d’Ulm, 75248 Paris Cedex 05, France
a r t i c l e i n f o
a b s t r a c t
Article history:
We investigated the structureeproperty relationship of liquid crystal elastomers (LCEs) obtained from a
series of nematic side-on monomers. A new synthetic strategy was developed to obtain the acrylate
monomers (n-ADBB), which gave us the opportunity to easily modify the spacer lengths of the mono-
mers. Through magnetic field alignment, well-defined LCE micropillars were fabricated from the
monomers by a method combining soft lithography and photopolymerization/photocrosslinking of the
monomers and a crosslinker. The influence of structural parameters on the thermomechanical defor-
mation of the microstructures was studied through microscopic observations. The study quantitatively
revealed the correlation of thermomechanical behavior of the microstructured LCEs with the crosslinking
density and length of the flexible spacer linking the mesogenic core to the backbone. With a proper
control of the structural parameters, optimized performances such as large reversible contraction, good
elasticity and mechanical robustness were demonstrated for this type of LCEs.
Received 12 March 2013
Received in revised form
28 June 2013
Accepted 21 July 2013
Available online xxx
Keywords:
Liquid crystal elastomers
Thermomechanical behavior
Molecular parameters
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
deformation of the elastomer sample. Making use of the sensitivity
of mesophases to various physical stimuli, LCEs with thermo-
Liquid crystalline elastomers (LCEs) have been intensively
investigated in recent years owing to their fascinating properties
and potential applications [1e10]. Combining anisotropic orienta-
tion of liquid crystals (LCs) with the rubbery elasticity of polymer
networks, the materials are endowed with specific properties, such
as stimuli-induced reversible deformation and anisotropic shape
changes. Depending on the LC phases exhibited by the systems,
LCEs are divided into nematic LCE [11], smectic LCE [8], and others.
The nematic LCE with uniform spatial orientation is the simplest
and most widely studied type among them. In a nematic LCE, the
average macromolecular shape is coupled with the orientational
nematic order [1,12]. The polymer chains elongate when their
mesogens orient in the nematic phase, while in the isotropic phase,
they recover a random coil conformation, driven by entropy [12]. At
the nematiceisotropic phase transition, a change in the average
molecular shape from elongated to coiled conformations will be
triggered and translated at a macroscopical level to induce a shape
responsive [13,14], photo-responsive [15,16], and electro-
responsive [17e19] functions have been developed. In recent
years, their applications as artificial muscles [20], micropumps and
microvalves for microfluidic devices [21e23], and opto-mechanical
shutters [24], have been actively explored.
Understanding structureeproperty relationship of the LCEs is
extremely important for optimizing properties and achieving real
applications. Different LCEs with the mesogenic units in the main-
chain and side-chains have been synthesized and investigated
[2,3,8,11,25e27]. For the LCEs with side-chain architecture, both
“end-on” and “side-on” LCEs, depending on whether the mesogenic
groups are attached terminally or laterally to the polymer backbone
via a flexible spacer, have been developed and intensively investi-
gated [2,3,8,11,25]. For liquid crystal polymers (LCPs) and LCEs, the
end-on structure usually tends to form a smectic phase, especially
for those with longer spacers [28]. On the other hand, the nematic
phase is favored in “side-on” LCPs and LCEs, in which the mesogens
are laterally attached to the polymer backbone [29e32]. In addition
to the mesogen-attaching mode, the mesomorphic properties of
LCPs and LCEs are also closely related to mesogenic units [33],
linking groups [34].
* Corresponding author. Tel.: þ86 10 62784561; fax: þ86 10 62770304.
** Corresponding author. Tel.: þ33 (0)1 56246762; fax: þ33 (0)1 40510636.
E-mail addresses: wxg-dce@mail.tsinghua.edu.cn (X. Wang), patrick.keller@
curie.fr (P. Keller).
There are several critically important issues need to be
addressed in order to further understand the structureeproperty
0032-3861/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.polymer.2013.07.057
Please cite this article in press as: Wei R, et al., Effect of molecular parameters on thermomechanical behavior of side-on nematic liquid crystal
elastomers, Polymer (2013), http://dx.doi.org/10.1016/j.polymer.2013.07.057

2
R. Wei et al. / Polymer xxx (2013) 1e9
relationship. The spacer length and crosslinking density are two
key control factors that need to be studied. The spacer length de-
termines the coupling strength between the side-chain mesogenic
units and the backbone, which plays an important role to influence
the molecular shape change during the phase transition. Cross-
linking density controls the topological restrain between the
polymeric chains and affects the mechanical properties related to
the chain conformation [35]. In order to explore the structuree
property relationship, nematic LCEs with uniaxial alignment and
mono-domain (named ‘liquid single crystal elastomer’ by Finkel-
mann [36]) are required. The ‘liquid single crystal elastomer’
samples have been obtained by methods such as stretching pre-
crosslinked films [2,4,8], drawing fibers from a polymer melt [9],
crosslinking in liquid crystalline cells [3,16], aligning with electric
or magnetic fields [13,14], and others [6,37e40]. Among them,
obtaining a regular array of pillar-like microstructures by soft-
lithography and magnetic field alignment is an appealing new
approach [13,14]. Although many efforts have been devoted to the
LCE study, to our knowledge, a systematic study on the effects of
the spacer length and crosslinking density by using the well-
defined microstructures has not been reported in the literature yet.
In this study, influences of crosslinking density and spacer
length on the properties of the side-on LCEs were investigated by
using pillar-like microstructures obtained by the method
combining soft-lithography and magnetic field alignment. In order
to keep the LCEs all in the nematic phase, the side-on nematic
monomers with different spacer length were selected. An improved
strategy was developed to synthesize side-on LC acrylate mono-
mers with the different spacer lengths. Monomers with different
spacer lengths, which contained 2, 3, 4, 5 and 6 methylenes (n-
ADBB, n representing the numbers of the methylenes), were syn-
thesized and characterized by 1H NMR, 13C NMR, FT-IR, POM and
DSC. The monomers were first mixed with the crosslinking agent in
different ratios and a suitable amount of the photoinitiator. Then,
LCE pillars were prepared with the mixture by the soft-lithographic
technique. The effects of the crosslinking density and spacer length
were studied by microscopically characterizing thermomechanical
properties of the LCE micropillars, which were separated from the
substrates. The correlation of thermomechanical behavior of the
micro-structured LCEs with the crosslinking density and length of
the flexible spacer was quantitatively revealed by the study.
spectrometer: the samples were mixed with KBr powder and then
pressed into thin transparent disks. The molecular weights and
molecular weight distributions were measured using a gel perme-
ation chromatographic (GPC) instrument equipped with a PLgel
5 mm mixed-D column and a refractive index (RI) detector (Wyatt
Optilab rEX). The measurements were carried out at 35 ꢀC and the
molecular weights were calibrated with polystyrene standards. THF
was used as the eluant and the flow rate was 1.0 mL/min. Thermal
analyses of the compounds were carried out using TA Instruments
DSC Q2000 system with a heating rate of 10 ꢀC/min in a nitrogen
atmosphere. Polarizing optical microscopic (POM) observations
were conducted on a Nikon LV 1000 POL microscope equipped with
a CCD camera and a hot stage. The SEM measurements were per-
formed on a field emission microscope (Hitachi S-4500) with the
accelerating voltage of 15 kV. The samples prepared for SEM studies
were observed after sputter coating treatment with Au.
2.3. 200-Acryloyloxyethyl 2,5-di(40-butyloxybenzoyloxy)benzoate
(2-ADBB)
The monomer was synthesized through three-step reactions as
shown in Scheme 1. 20-Bromo-ethyl 2,5-dihydroxybenzoate (A2)
was first obtained from the nucleophilic substitution reaction be-
tween 2,5-dihydroxybenzoic acid and 1,2-dibromoethane. Then, 20-
acryloyloxyethyl 2,5-dihydroxybenzoate (B2) was obtained by re-
action between A2 and acrylic acid. The synthetic details and ana-
lytic results of the intermediates are given in the Supporting
information. In the final step, a solution of B2 (2.5 g, 10 mmol), 4-
n-butyloxybenzoic acid (4.1 g, 21 mmol), N,N-dicyclohex-
ylcarbodiimide (6.2 g, 30 mmol), and 4-pyrrolidinopyridine (0.48 g,
3 mmol) in dichloromethane (100 mL) was stirred at room tem-
perature for 24 h. The N,N-dicyclohexyl urea was filtered and the
filtrate was sequentially washed with water (150 mL), 5% acetic acid
solution (150 mL), and water (150 mL), and dried over MgSO4. After
evaporation of the solvent, the residue was subjected to column
chromatography on silica gel with DCM as eluting solvent to yield
white powder (70%). 1H NMR (600 MHz, d6-DMSO)
d (ppm): 8.10
(m, 4H, ArH), 7.83 (d, 1H, ArH), 7.65 (m, 1H, ArH), 7.50 (d, 1H, ArH),
7.13 (m, 4H, ArH), 6.27, 6.07, 5.88 (3m, 3H, CH2]CH), 4.38, 4.20 (2t,
4H, eCH2eO), 4.11 (t, 4H, eCH2eO),1.74,1.46 (2m, 8H, eCH2e), 0.96
(t, 6H, eCH3). 13C NMR (150 MHz, d6-DMSO)
d (ppm): 14.2, 19.2,
31.1, 62.4, 63.6, 68.2, 68.3, 115.2, 115.3, 121.3, 121.5, 124.3, 124.5,
126.0, 128.4, 128.6, 131.9, 132.3, 132.6, 132.7, 148.1, 148.3, 163.7,
163.9, 164.8 and 165.8. IR (KBr, cmꢁ1): 3078 (eC]CeH, s), 2956,
2936, 2870 (CeH, s), 1731 (C]O, s), 1609, 1582, 1512 (Benz. ring, s),
2. Experimental section
2.1. Materials
1470 (CeH, d), 1302, 1250, 1200 1166 (CeOeC, s). EA: C 68.0 (calcd
2,5-Dihydroxybenzoic acid (99%) and 4-n-butyloxybenzoic acid
(98%) were purchased from Alfa Aesar and used as received.
Crosslinking agent, 1,6-hexanediol diacrylate (98%) was purchased
from Adamas-beta and used as received. Photoinitiator, 2-benzyl-2-
(dimethylamino)-40-morpholinobutyrophenone (97%), was pur-
chased from Sigma Aldrich and used as received. Tetrahydrofuran
(THF) was purified by distillation with sodium and benzophenone.
Deionized water (resistivity > 18 MU cm) was obtained from a
Milli-Q water purification system. Azo-bis-isobutryonitrile (AIBN)
was recrystallized from anhydrous methanol before use. All other
reagents were commercially available products and used as
received without further purification.
67.5), H 5.9 (calcd 5.9).
2.4. 300-Acryloyloxypropyl 2,5-di(40-butyloxybenzoyloxy)benzoate
(3-ADBB)
3-ADBB was similarly prepared as mentioned for the 2-ADBB
synthesis. The synthetic details and analytic results of the in-
termediates A3 and B3 are given in the Supporting information. 1H
NMR (300 MHz, d6-DMSO)
d (ppm): 8.10 (m, 4H, ArH), 7.85 (d, 1H,
ArH), 7.65 (m, 1H, ArH), 7.49 (d, 1H, ArH), 7.12 (m, 4H, ArH), 6.24,
6.10, 5.88 (3m, 3H, CH2]CH), 4.20 (t, 2H, eCH2eO), 4.10 (m, 6H, e
CH2eO), 1.75 (m, 6H, eCH2e), 1.47 (m, 4H, eCH2e), 0.95 (t, 6H, e
CH3). 13C NMR (75 MHz, d6-DMSO)
d (ppm): 14.2, 19.2, 27.8, 31.1,
2.2. Characterization
61.5, 62.2, 68.3, 115.3, 120.9, 121.2, 124.8, 125.0, 126.1, 128.5, 128.7,
132.0, 132.7, 148.1, 148.7, 163.8, 164.2, 164.8 and 165.8. IR (KBr,
cmꢁ1): 3077 (eC]CeH, s), 2958, 2934, 2869 (CeH, s), 1731 (C]O,
s), 1634 (C]C, s), 1609, 1582, 1512 (Benz. ring, s), 1475, 1392 (CeH,
1H and 13C NMR spectra were obtained on a JEOL JNM-ECA300 or
JEOL JNM-ECA600 NMR spectrometer with tetramethylsilane (TMS)
as the internal standard at ambient temperature in d6-DMSO or
d), 1251, 1182, 1166, 1075 (CeOeC, s). EA : C 67.9 (calcd 68.0), H 6.1
CDCl3. FT-IR spectra were collected on
a Nicolet 560-IR
(calcd 6.1).
Please cite this article in press as: Wei R, et al., Effect of molecular parameters on thermomechanical behavior of side-on nematic liquid crystal
elastomers, Polymer (2013), http://dx.doi.org/10.1016/j.polymer.2013.07.057

R. Wei et al. / Polymer xxx (2013) 1e9
3
Scheme 1. Synthetic route of the monomers.
2.5. 400-Acryloyloxybutyl 2,5-di(40-butyloxybenzoyloxy)benzoate
(4-ADBB)
14.2, 19.2, 22.3, 28.0, 28.2, 31.1, 64.3, 65.6, 68.3, 68.4, 115.2, 120.9,
121.1, 125.1, 126.1, 128.8, 129.0, 132.5, 132.7, 147.9, 148.6, 163.8, 164.1,
164.6 and 166.0. IR (KBr, cmꢁ1): 3063 (eC]CeH, s), 2958, 2934,
2872 (CeH, s), 1727, 1704 (C]O, s), 1634 (C]C, s), 1604, 1578, 1511
(Benz. ring, s), 1488, 1389 (CeH, d), 1252, 1163, 1065 (CeOeC, s). EA:
C 68.7 (calcd 68.7), H 6.5 (calcd 6.5).
4-ADBB was similarly prepared as mentioned for the 2-ADBB
synthesis. The synthetic details and analytic results for the in-
termediates A4 and B4 are given in the Supporting information. 1H
NMR (300 MHz, d6-DMSO)
d (ppm): 8.09 (m, 4H, ArH), 7.84 (d, 1H,
2.7. 600-Acryloyloxyhexyl 2,5-di(40-butyloxybenzoyloxy)benzoate
(6-ADBB)
ArH), 7.66 (m, 1H, ArH), 7.49 (d, 1H, ArH), 7.12 (m, 4H, ArH), 6.25,
6.13, 5.91 (3m, 3H, CH2]CH), 4.11 (m, 6H, eCH2eO), 3.95 (t, 2H, e
CH2eO), 1.72 (m, 4H, eCH2e), 1.48 (m, 8H, eCH2e), 0.95 (t, 6H,
CH3). 13C NMR (75 MHz, d6-DMSO)
d (ppm): 15.3, 20.3, 26.2, 32.2,
6-ADBB was similarly prepared as mentioned for the 2-ADBB
synthesis. The synthetic details and analytic results for the in-
termediates A6 and B6 are given in the Supporting information. 1H
65.2, 66.5, 69.4, 116.4, 122.3, 126.1, 126.8, 129.5, 130.0, 132.9, 133.9,
148.8, 149.4, 164.9, 165.2, 165.9 and 166.4. IR (KBr, cmꢁ1): 3075 (e
C]CeH, s), 2958, 2934, 2872 (CeH, s), 1732 (C]O, s), 1634 (C]C,
s), 1608, 1581, 1512 (Benz. ring, s), 1475, 1389 (CeH, d), 1252, 1182,
1166, 1073 (CeOeC, s). EA: C 68.2 (calcd 68.3), H 6.3 (calcd 6.3).
NMR (300 MHz, CDCl3) d (ppm): 8.15 (m, 4H, ArH), 7.88 (d,1H, ArH),
7.46 (m,1H, ArH), 7.26 (d,1H, ArH), 6.98 (m, 4H, ArH), 6.35, 6.13, 5.80
(3m, 3H, CH2]CH), 4.09 (m, 8H, eCH2eO),1.81 (m, 4H, eCH2e),1.55
(m, 8H, eCH2e), 1.27 (m, 4H, eCH2e), 0.99 (t, 6H, eCH3). 13C NMR
2.6. 500-Acryloyloxypentyl 2,5-di(40-butyloxybenzoyloxy)benzoate
(5-ADBB)
(75 MHz, CDCl3) d (ppm): 13.9, 19.3, 25.6, 28.4, 28.5, 31.2, 53.5, 64.5,
65.5, 68.1, 114.4, 114.5, 121.1, 121.5, 125.1, 127.2, 128.7, 130.5, 132.4,
132.5,148.1,148.4,163.7,164.2,164.7 and 166.3. IR (KBr, cmꢁ1): 3063
(eC]CeH, s), 2957, 2935, 2871 (CeH, s), 1732 (C]O, s), 1634 (C]C,
s), 1608, 1581, 1513, 1480 (Benz. ring, s), 1492, 1389 (CeH, d), 1251,
1166, 1073 (CeOeC, s). EA: C 69.0 (calcd 69.1), H 6.7 (calcd 6.7).
5-ADBB was similarly prepared as mentioned for the 2-ADBB
synthesis. The synthetic details and analytic results for the in-
termediates A5 and B5 are given in the Supporting information. 1H
NMR (600 MHz, d6-DMSO)
d (ppm): 8.09 (m, 4H, ArH), 7.83 (d, 1H,
ArH), 7.66 (m, 1H, ArH), 7.47 (d, 1H, ArH), 7.13 (m, 4H, ArH), 6.29,
6.13, 5.89 (3m, 3H, CH2]CH), 4.11 (m, 6H, eCH2eO), 3.98 (t, 2H, e
CH2eO), 1.74 (m, 4H, eCH2e), 1.45 (m, 8H, eCH2e), 1.22 (m, 2H, e
2.8. Polymerization of the monomers in solution
To test the polymerization ability, the monomers were also
polymerized in solutions. Description of the radical polymerization
CH2e), 0.96 (t, 6H, eCH3). 13C NMR (150 MHz, d6-DMSO)
d (ppm):
Fig. 1. DSC curves of the monomers: a. DSC curves of the monomers in the first heating scan, from bottom to top: 2-ADBB, 3-ADBB, 4-ADBB, 5-ADBB and 6-ADBB; b. DSC curves of
the monomers in the first cooling scan, from bottom to top: 2-ADBB, 3-ADBB, 4-ADBB, 5-ADBB and 6-ADBB.
Please cite this article in press as: Wei R, et al., Effect of molecular parameters on thermomechanical behavior of side-on nematic liquid crystal
elastomers, Polymer (2013), http://dx.doi.org/10.1016/j.polymer.2013.07.057

4
R. Wei et al. / Polymer xxx (2013) 1e9
Table 1
gently pressed down on the melted sample, which filled the inner
structure of the mold. The temperature was then slowly decreased
(1 ꢀC/min) down to 60 ꢀC, at which the sample is in its nematic
phase. For 6-ADBB, the temperature was decreased down to 45 ꢀC
due to the fact that the TNI is lower than 60 ꢀC. Then the sample was
irradiated with the UV light from a high pressure mercury lamp
(1000 W, equipped with quartz jacket condenser cooled with
running water) for 1 h for the polymerization and crosslinking. The
distance between the sample and lamp was about 25 cm. The
whole setup was shielded by quartz cover through which argon
flow could get through to ensure an inert atmosphere above the
sample. The argon gas protection was kept until the polymerization
and crosslinking were completed. After cooling to room tempera-
ture, the PDMS mold was peeled off, leaving a thin glassy polymer
film covered by a regular array of pillars. Separate pillars were
obtained by carefully cutting them off from the substrate with a
razor blade and suspending them in silicon oil.
The TNI of the monomers obtained from DSC and POM observation; TNI of the LCE
micropillars estimated from POM observation, where the crosslinking agent content
was 20 mol%.
TNI ꢀC (DSC)
monomer
TNI ꢀC (POM)
monomer
TNI ꢀC (POM)
LCE pillars
2-ADBB
3-ADBB
4-ADBB
5-ADBB
6-ADBB
100
92
87
82
79
104
95
88
85
81
110
115
125
116
115
in solution of the monomers and characterization of the corre-
sponding side-on polymers are given in the Supporting
information.
2.9. Preparation of PDMS molds
3. Results and discussion
Photoresist was spin-coated on a piece of pre-cleaned silicon
wafers and dried in vacuum. The thickness of the photoresist was
3.1. Synthesis and characterization
set to about 100
mm by controlling the spinning speed. A photo-
mask with 4 ¼ 20
m
m hole array was placed in contact with the
The synthetic route used to prepare the monomers (n-ADBB) is
shown in Scheme 1, which includes three reaction steps. Firstly, a
nucleophilic substitution reaction between the carboxylate group of
surface of the photoresist. After illumination with ultraviolet (UV)
light through the mask, the photoresist not crosslinked was dis-
solved and removed by an organic solvent. The master with pillar
array of photoresist in bas-relief on the surface was obtained. The
poly(dimethylsiloxane) (PDMS) molds were obtained by mixing the
elastomer base and curing agent (Dow Corning, Sylgard 184) in a
proper ratio (10:1, w/w). The prepolymer was poured and casted on
the master surface and then cured in a 40 ꢀC oven for 4 h. After
curing, the PDMS molds were separated from the master and used
afterwards without any more treatment.
2,5-dihydroxybenzoic acid and one of the bromine in a,u-dibro-
moalkanes was carried out under mild conditions. Then, the second
bromine was substituted with the reactive acrylate group using the
same procedure. Finally, the monomers were obtained by esterifi-
cation of the previous intermediates with 4-n-butyloxybenzoic acid
under standard conditions (activated by dicyclohexylcarbodiimide
and catalyzed by 4-pyrrolidinopyridine). This newly developed
procedure is simple and straightforward compared with the previ-
ously reported scheme [11,41]. It allowed us to efficiently synthesize
monomers n-ADBB with different spacer lengths (n ¼ 2, 3, 4, 5, 6).
The analytical results, obtained from 1H NMR, 13C NMR, and FT-IR,
are given in the Experimental section and 1H NMR spectra with
resonance signal assignments are given in the Supporting
information. The results confirmed that the series of the nematic
side-on acrylate monomers were successfully synthesized.
2.10. Fabrication of pillars
Monomers and crosslinking agent 1,6-hexanediol diacrylate in
different ratios were mixed with 2% photoinitiator 2-benzyl-2-
(dimethylamino)-40-morpholinobutyrophenone, using dichloro-
methane as a solvent. According to the literature [13,14], a small
amount of the mixture was heated to the isotropic phase (100 ꢀC)
on a microscope slide positioned atop a rare earth permanent
magnet (1 T NdFeB rare earth magnet). The PDMS mold was then
The mesomorphic properties of the monomers were charac-
terized by DSC and POM. Fig. 1 shows the DSC curves of the
monomers on the first heating (Fig. 1a) and first cooling (Fig. 1b)
Fig. 2. POM images of the monomers. a. 2-ADBB at 103 ꢀC; b. 3-ADBB at 94 ꢀC; c. 4-ADBB at 87 ꢀC; d. 5-ADBB at 84 ꢀC; e. 6-ADBB at 65 ꢀC; f. 6-ADBB at 80 ꢀC.
Please cite this article in press as: Wei R, et al., Effect of molecular parameters on thermomechanical behavior of side-on nematic liquid crystal
elastomers, Polymer (2013), http://dx.doi.org/10.1016/j.polymer.2013.07.057

R. Wei et al. / Polymer xxx (2013) 1e9
5
Fig. 3. Optical micrographs and SEM images of the LCE micropillars: a. optical microscopic graphs of the array of the LCE pillars; b. separate LCE pillars suspended in silicon oil; c.
and d. SEM images of LCE pillars. The LCE pillars were obtained from 4-ADBB with the 20 mol% crosslinking agent.
scans. 2-ADBB melts at 96 ꢀC (peak value) and shows a small
endothermic transition at 100 ꢀC (peak value), which corresponds
to the clearing point (TNI) of the monomer on heating. On the
cooling scan, an exothermic transition around 100 ꢀC (peak value) is
observed for 2-ADBB, while no crystallization peak can be detected.
The other monomers show lower melting and clearing points as the
spacer length increases (Fig. 1 and Table 1). POM observations
confirm that all the monomers enter the nematic phase after
melting. Fig. 2 shows some representative images of the textures
observed by POM.
In order to compare with the properties of the LCEs, the
monomers were also polymerized by radical polymerization in
solution. The characterization of the corresponding side-on liquid
crystalline polymers gives further information on the physical
properties in particular their mesomorphic properties. The ob-
tained polymers have average molecular weights (Mn) around
10,000 and polydispersities (PDI) below 2.5 (Table S1, in the
Supporting information). The phase transition temperatures ob-
tained by both DSC and POM observation are also given there. All
the polymers exhibit a nematic mesophase in proper temperature
ranges. The nematic-isotropic transition temperatures (TNI) of the
polymers decrease with the spacer length increase.
3.2. Fabrication of micropillar arrays and separate micropillars
Using the nematic acrylate monomers synthesized above, LCEs
were prepared under the form of thin films covered with micro-
pillars. The method was described in the Experimental section and
also in the previous reports in detail [13,14]. In the process,
monomers and crosslinking agent 1,6-hexanediol diacrylate in
different ratios were mixed with 2% (mol%) of the photoinitiator, 2-
benzyl-2-(dimethylamino)-40-morpholinobutyrophenone.
After
heated the mixture to the isotropic phase, a piece of the PDMS mold
was gently pressed down on the melted sample, which filled the
inner column-shaped structures of the mold. When the sample was
cooled down to enter the nematic phase with the magnetic field
alignment, it was irradiated by the UV light with the argon gas
protection for the polymerization and crosslinking. A thin glassy
polymer film covered by a regular array of pillars was obtained after
cooling to room temperature and peeling off the PDMS mold.
Fig. 4. POM images of micropillars obtained from 4-ADBB with the 5 mol% crosslinking agent, observed at room temperature.
Please cite this article in press as: Wei R, et al., Effect of molecular parameters on thermomechanical behavior of side-on nematic liquid crystal
elastomers, Polymer (2013), http://dx.doi.org/10.1016/j.polymer.2013.07.057

6
R. Wei et al. / Polymer xxx (2013) 1e9
Fig. 5. The optical micrographs of a single LCE pillar at different temperatures: a. room temperature; b. 130 ꢀC; c. room temperature again. The LCE pillar was obtained from 4-ADBB
with 15 mol% crosslinking agent.
Separate micropillars were obtained from the pillar array by care-
fully cutting them off from the substrate with a razor blade.
Fig. 3 shows the optical microscope and SEM images of the
micropillar array and separate pillars. In Fig. 3a, the optical mi-
croscope image shows regular alignment of the micropillars in the
array on a large scale. It indicates that micro-structures with uni-
form sizes can be obtained with a high efficiency by this method.
Fig. 3b shows the optical microscope image of the separate pillars
suspended in silicon oil. The average length of the micropillars was
the temperature with a slow temperature variation (with the heat-
ing and cooling rates ꢂ1 ꢀC per minute) from 100 ꢀC to 130 ꢀC and
back. Fig. 6 shows the deformation of the micropillars with the
temperature, where the pillars were made from 4-ADBB and
different contents of the crosslinking agent. The deformation in the
figure is given as the length ratio between the contracted pillar and
the original one, and the content of crosslinking agent is presented
by the molar percentage. All the LCE micropillars show significant
contraction except the one with 25% crosslinking agent. It can be
attributed to the high crosslinking density in the system. With high
concentration of the crosslinking agent, too many crosslinking
junctions are introduced, which causes the sample to change from
an elastomer to a thermoset. It is interesting to note that a sudden
change occurs when the crosslinking agent increases from 20% to
25%. This process can be rationalized by considering a percolation
change with the crosslinking density. Among the LCE micropillars
showing the thermomechanical responses, the one with 5% cross-
linking agent exhibits the largest contraction, reflecting the fact that
it is a softer elastomer. The measured contractions are 39%, 30%, 27%,
24% and 0% (the contraction is defined as (L0 ꢁ LT)/L0) for micropillars
with crosslinking agent content of 5%, 10%, 15%, 20% and 25%
respectively. The temperatures for the LCE micropillars to reach the
maximum contraction decrease with the increasing crosslinking
agent contents in a few-degree scale. The LCE micropillar with 5%
crosslinking agent shows the highest transition temperature among
them. This observation can be attributed to the flexible and non-
mesogenic nature of the crosslinking agent, where the increase in
measured to be around 62 mm Fig. 3c and 3d shows the SEM images
of the pillars for different amplification scales. The pillar sizes were
also measured from the SEM images by carefully cutting off the
pillars before the sputtering treatment. The lengths of the micro-
pillars were estimated to be about 60 mm. For all five monomers
prepared in this work, the well-defined pillar-like microstructures
were obtained and used for the following investigations. Fig. 4 gives
some typical POM images of the separate micropillars placed be-
tween the crossed polarizers and observed at the room tempera-
ture. The bright images due to the birefringence evidence the
mesogenic alignment in the micropillars. Similar optical anisotropy
is observed for the micropillars obtained from the other monomers
with different spacer lengths.
3.3. Effect of the crosslinking density
The effect of the degree of crosslinking on the thermomechanical
behaviorof the LCEswasstudied byusing themicropillars. Theresult
obtained for 4-ADBB is presented belowto elucidate the effect. In the
experiment, 4-ADBB and 2 mol% photoinitiator were mixed with 5,
10, 15, 20 and 25 mol% of crosslinking agent 1,6-hexanediol dia-
crylate, respectively. The regular arrays of LCE micropillars on glass
substrates and separate micropillars were obtained from these
samples by the method mentioned above. The shape changes of
these individual micropillars, suspended in silicon oil to prevent
them from sticking on the microscope glass slides, were studied as a
function of the temperature by POM equipped with the hot stage.
Fig. 5 shows typical optical micrographs of a single LCE micro-
pillar obtained from 4-ADBB with 15 mol% of the crosslinking agent
at room temperature (Fig. 5a and c) and 130 ꢀC (Fig. 5b). By heating
to a temperature above the nematic-to-isotropic phase transition
temperature of the LCE, a significant contraction of the micropillar
in the longitudinal direction is observed. While cooling back to
room temperature, the micropillar recovers its original shape. As
the micropillars prepared by this method are uniform, the ther-
momechanical deformation in the same scale is also observed for
other pillars in the array. It can be estimated from the micrographs
that the micropillar contracts up to 77% of the original length in the
longitudinal direction.
its content will cause the decrease of orientational order and TNI
Above result indicates that 5% of the crosslinking agent seems to
be sufficient to transform a visco-elastic polymer melt into an
.
To better understand the thermomechanical properties of the
LCEs, the contraction and relaxation were measured as a function of
Fig. 6. The length change of the LCE pillars with different contents of crosslinking
agent versus the changing temperature. The LCE pillars were obtained from 4-ADBB.
Please cite this article in press as: Wei R, et al., Effect of molecular parameters on thermomechanical behavior of side-on nematic liquid crystal
elastomers, Polymer (2013), http://dx.doi.org/10.1016/j.polymer.2013.07.057

R. Wei et al. / Polymer xxx (2013) 1e9
7
contraction becomes smaller as the spacer length of the monomer
increases (Fig. 8a). As the spacer lengths increase, the contraction
gradually becomes smaller when the spacer length increases from
2-ADBB to 6-ADBB. The result shows that the coupling between the
mesogenic units and backbone is important to force the polymer
chain to take an anisotropic conformation. As the short spacer in 2-
ADBB forces the backbone to take a more anisotropic shape, which
acts more or less like a main-chain liquid crystal polymer, a larger
contraction will occur at the nematic to isotropic phase transition.
When the spacer length becomes longer, the coupling between the
mesogens and backbone through the spacers becomes weaker. The
monomer possessing a long spacer, such as 6-ADBB, could behave
more similar to an end-on side-chain liquid crystal polymer. For all
samples, when the temperature gradually decreases, the micro-
pillars show the reversible expansion as shown in Fig. 8b.
Another remarkable point that can be seen from Fig. 8 is that the
transition temperature of the LCEs is also closely related to the
spacer length. However, the result shows a somewhat irregular
variation of the transition temperatures with the spacer length
increase. The micropillars obtained from 4-ADBB show the highest
transition temperature compared with others. To understand this
phenomenon, the TNI values of LCE micropillars were obtained from
the POM observation and compared with the temperatures for the
micropillars to reach the maximum contraction. The TNI of LCE
micropillars obtained from the POM observation is given in Table 1
together with the transition temperatures of the corresponding
monomers. The TNI values are 110, 115, 125, 116, and 115 ꢀC for LCE
micropillars obtained from 2-ADBB, 3-ADBB, 4-ADBB, 5-ADBB, and
6-ADBB. Compared with the thermomechanical deformation
behavior shown in Fig. 8a, it can be seen that for the LCE micro-
pillars obtained from 2-ADBB, 3-ADBB, and 5-ADBB, the tempera-
ture to reach the maximum contraction is about the 5e10 ꢀC lower
compared to the corresponding TNI values. On the contrary, the
temperatures for LCE micropillars obtained from 4-ADBB and 6-
ADBB to reach the maximum contraction are almost the same as
Fig. 7. The reversible switch between contraction and extension of the LCE pillars
obtained from 4-ADBB.
elastomer. To further confirm this point, the thermomechanical
deformation behavior was investigated through repeated contrac-
tioneextension variations by cycling up to ten times. The good
reversibility and robustness shown in Fig. 7 verify the elasticity of
the materials. This result not only shows the potential applicability
of this kind of LCE materials but also indicates that the addition of
the 5% crosslinking agent is enough to introduce the crosslinking
density above the threshold value.
3.4. Effect of the spacer length
The result obtained from 4-ADBB as the representative shows
that LCE microstructure with reversible thermomechanical defor-
mation can be obtained by introducing a proper amount of cross-
linking points. By similar method, LCE micropillars were prepared
from the other synthesized acrylate monomers to study the effect
of the spacer length on the properties. In all experiments reported
below, the concentration of the crosslinking agent was fixed to be
20 mol%. The results observed on samples with concentrations of
the crosslinking agent of 5, 10, 15 mol% are similar.
Fig. 8 shows the contraction and expansion of the micropillars
made from this series of monomers with the temperature variation.
The deformation in the figure is given as the length ratio of the
contracted pillar to the original one. The micropillars made from 2-
ADBB exhibit the largest contraction of about 31% and the
the corresponding TNI
.
The observations indicate that the relationship between the
thermomechanical transition temperature and spacer length can be
attributed to the two contradictory factors that influence the
transition temperature. As indicated by the result given in Fig.1 and
Table 1, the clearing temperatures of the monomers decrease with
the increasing spacer length. Similar relationship can also be seen
for corresponding linear polymers obtained from the monomers
through the radical polymerization (Table S1, in the Supporting
information). Therefore, the transition temperature of LCEs
should show a similar tendency with the spacer length variation.
Fig. 8. The length change of the LCE pillars made from the monomers: a, contraction, b, expansion. The content of the crosslinking agent was 20 mol%.
Please cite this article in press as: Wei R, et al., Effect of molecular parameters on thermomechanical behavior of side-on nematic liquid crystal
elastomers, Polymer (2013), http://dx.doi.org/10.1016/j.polymer.2013.07.057

8
R. Wei et al. / Polymer xxx (2013) 1e9
The crosslinking density of the LCEs showed a significant in-
fluence on the thermomechanical properties of LCEs. If the density
of the crosslinking points was too high, such as the system with 25%
crosslinking agent, no thermal-induced contraction could be
observed. The observed contractions changed from 39%, 30%, 27% to
24% for the micropillars with the crosslinking agent content of 5%,
10%, 15% and 20%, respectively. The 5 mol% amount of the cross-
linking agent was sufficient to transform a visco-elastic polymer
melt into an elastomer.
The spacer length of the LCEs was another important factor to
influence the thermomechanical properties of LCEs. For LCEs with
the short spacer, the coupling between the mesogens and polymer
main-chain was strong, which induced a more anisotropic chain
conformation. The micropillars made from 2-ADBB showed the
largest contraction of about 31% and the contraction decreased as
the spacer length in the monomers became longer. For the systems
with a proper amount of the crosslinking agent, the spacer length
variation from 2 to 6 methylenes showed no influence on the
elasticity.
Fig. 9. The reversible switch between contraction and extension of the LCE pillars
obtained from the monomers. The content of crosslinking agent was 20 mol%.
Acknowledgment
This tendency can be seen for LCEs obtained from 4-ADBB, 5-ADBB,
and 6-ADBB. On the other hand, LCE with the short spacer has the
strong correlation between the mesogens and backbone. As a
result, the polymeric chains between the crosslinks will be dis-
torted more significantly with respect to the equilibrium confor-
mation. This departure from the equilibrium can be treated as an
internal tension, which will trigger the relaxation to occur at a
lower temperature to maximize the entropy. This internal tension is
the cause for LCEs obtained from 2-ADBB, 3-ADBB and 5-ADBB to
reach the maximum contraction at the temperatures about the 5e
10 ꢀC below the corresponding TNI values. As another evidence of
this internal tension, it can be seen from the data in Table 1 that the
LCE pillars obtained from 2-ADBB show the smallest increase in TNI
after the polymerization and crosslinking. The LCE pillars obtained
from 4-ADBB show the largest increase in TNI after the polymeri-
zation and crosslinking. Therefore, the TNI values for LCEs obtained
from 2-ADBB and 3-ADBB are actually lower than that of LCE ob-
tained from 4-ADBB. This understanding can be used to explain
why the micropillars obtained from 2-ADBB and 3-ADBB show the
lower thermomechanical transition temperatures compared with
that of 4-ADBB. The lower thermomechanical transition tempera-
tures of LCEs obtained from 5-ADBB and 6-ADBB can be attributed
to the lower TNI for the monomers and polymers. However, due to
the internal tension, TNI and the thermomechanical transition
temperature of LCE obtained from 5-ADBB appear to be lower than
that it should be.
The financial support from the NSFC under Projects 91027024,
51233002 and 51061130556, from the French ANR under Project
ANR-10-INTB-0904 and
acknowledged.
a
PICS-CNRS project is gratefully
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.polymer.2013.07.057.
References
[1] Warner M, Terentjev EM. Liquid crystal elastomers. Oxford, UK: Oxford Uni-
versity Press; 2003.
[2] Finkelmann H, Kim ST, Munoz A, Palffy-Muhoray P, Taheri B. Adv Mater
2001;13(14):1069e72.
[3] Ikeda T, Nakano M, Yu YL, Tsutsumi O, Kanazawa A. Adv Mater 2003;15(3):
201e5.
[4] Ji Y, Huang YY, Rungsawang R, Terentjev EM. Adv Mater 2010;22(31):3436e40.
[5] Li MH, Keller P, Li B, Wang XG, Brunet M. Adv Mater 2003;15(7e8):569e72.
[6] Ohm C, Serra C, Zentel R. Adv Mater 2009;21(47):4859e62.
[7] van Oosten CL, Bastiaansen Cees WM, Broer DJ. Nat Mater 2009;8(8):677e82.
[8] Lehmann W, Skupin H, Tolksdorf C, Gebhard E, Zentel R, Krüger P, et al. Nature
2001;410(6827):447e50.
[9] Yoshino T, Kondo M, Mamiya JI, Kinoshita M, Yu YL, Ikeda T. Adv Mater
2010;22(12):1361e3.
[10] Yamada M, Kondo M, Mamiya JI, Yu YL, Kinoshita M, Barrett CJ, et al. Angew
Chem Int Ed 2008;47(27):4986e8.
[11] Thomsen DL, Keller P, Naciri J, Pink R, Jeon H, Shenoy D, et al. Macromolecules
2001;34(17):5868e75.
[12] Cotton JP, Hardouin F. Prog Polym Sci 1997;22(4):795e828.
[13] Buguin A, Li MH, Silberzan P, Ladoux B, Keller P. J Am Chem Soc 2006;128(4):
1088e9.
[14] Yang H, Buguin A, Taulemesse JM, Kaneko K, Méry S, Bergeret A, et al. J Am
Chem Soc 2009;131(41):15000e4.
[15] Finkelmann H, Nishikawa E, Pereira GG, Warner M. Phys Rev Lett 2001;87(1).
015501, 4 pages.
[16] Yu YL, Nakano M, Ikeda T. Nature 2003;425(6954):145.
[17] CourtyS, Mine J, Tajbakhsh AR, Terentjev EM. Europhys Lett2003;64(5):654e60.
[18] Huang C, Zhang QM, Jákli A. Adv Funct Mater 2003;13(7):525e9.
[19] Urayama K, Honda S, Takigawa T. Macromolecules 2005;38(9):3574e6.
[20] Ohm C, Brehmer M, Zentel R. Adv Polym Sci 2012;250:49e94.
[21] Chen ML, Xing X, Liu Z, Zhu YT, Liu H, Yu YL, et al. Appl Phys A-Mater
2010;100(1):39e43.
[22] Chen ML, Huang HT, Zhu YT, Liu Z, Xing X, Cheng FT, et al. Appl Phys A-Mater
2011;102(3):667e72.
[23] Fleischmann1 EK, Liang HL, Kapernaum N, Giesselmann F, Lagerwall J,
Zentel R. Nat Commun 2012;3:1178e85.
Similar to the result given in Fig. 7, a good reversibility and
robustness can also be observed for the LCEs with the different
spacer lengths. Fig. 9 shows the repeated contractioneextension
variations by cycling up to ten times, which verifies the good
elasticity of the materials. Above result indicates that for the sys-
tems with the proper amount of the crosslinking agent, the spacer
length variation from 2 to 6 methylenes shows no influence on
transformation from visco-elastic to elastic after the crosslinking.
4. Conclusions
In order to investigate the structureeproperty relationship, we
synthesized a series of nematic side-on liquid crystal acrylate
monomers (n-ADBB) by a modified synthetic strategy. The mono-
mers were characterized by 1H NMR, 13C NMR, FT-IR, POM and
DSC. LCE micropillars were made from these newly synthesized
monomers by the soft lithography and photopolymerization/
photocrosslinking.
[24] Serak S, Tabiryan N, Vergara R, White TJ, Vaia RA, Bunning TJ. Soft Matter
2010;6(4):779e83.
[25] Ogawa H, Stibal-Fischer E, Finkelmann H. Macromol Chem Phys 2004;205(5):
593e9.
Please cite this article in press as: Wei R, et al., Effect of molecular parameters on thermomechanical behavior of side-on nematic liquid crystal
elastomers, Polymer (2013), http://dx.doi.org/10.1016/j.polymer.2013.07.057

R. Wei et al. / Polymer xxx (2013) 1e9
9
[26] Beyer P, Braun L, Zentel R. Macromol Chem Phys 2007;208(22):2439e48.
[27] Bispo M, Guillon D, Donnio B, Finkelmann H. Macromolecules 2008;41(9):
3098e108.
[28] Finkelmann H, Rehage G. Adv Polym Sci 1984;60/61:99e172.
[29] Gray GW, Hill JS, Lacey D. Mol Cryst Liq Cryst 1991;197(1):43e55.
[30] Mu J, Okamoto H, Yanai T, Takenaka S, Feng XS. Colloid Surf A 2001;181(1e3):
303e13.
[31] Lecommandoux S, Achard MF, Hardouin F. Liq Cryst 1998;25(1):85e94.
[32] Achard MF, Lecommandoux S, Hardouin F. Liq Cryst 1995;19(5):581e7.
[33] (a) Ober CK, Jin J, Zhou QF, Lenz RW. Adv Polym Sci 1984;59:103e46;
(b) Chen S, Luo H, Xie HL, Zhang HL. Polymer 2013;54(7):1794e802.
[34] Chen XF, Shen ZH, Wan XH, Fan XH, Chen EQ, Ma YG, et al. Chem Soc Rev
2010;39(8):3071e101.
[35] Yu YL, Nakano M, Shishido A, Shiono T, Ikeda T. Chem Mater 2004;16(9):
1637e43.
[36] Küpfer J, Finkelmann H. Macromol Rapid Commun 1991;12(12):717e26.
[37] Ohm C, Brehmer M, Zentel R. Adv Mater 2010;22(31):3366e87.
[38] Ohm C, Kapernaum N, Nonnenmacher D, Giesselmann F, Serra C, Zentel R.
J Am Chem Soc 2011;133(14):5305e11.
[39] Ohm C, Haberkorn N, Theato P, Zentel R. Small 2011;7(2):194e8.
[40] Wei RB, He YN, Wang XG, Keller P. Macromol Rapid Commun 2013;34(4):330e4.
[41] Keller P, Thomsen DL, Li MH. Macromolecules 2002;35(2):581e4.
Please cite this article in press as: Wei R, et al., Effect of molecular parameters on thermomechanical behavior of side-on nematic liquid crystal
elastomers, Polymer (2013), http://dx.doi.org/10.1016/j.polymer.2013.07.057