Vol. 3

LIQUID CRYSTALLINE THERMOSETS

139

LIQUID CRYSTALLINE THERMOSETS

Introduction

Liquid crystalline thermosets (LCTs) may generally be defined as low molar mass,
multifunctional monomers, which can be cured thermally, chemically, or photo-
chemically in the melt state, leading to a highly cross-linked, high glass-transition
temperature material which exhibits liquid crystalline order. These materials are
expected to exhibit properties that combine the useful benefits of both cross-linked
thermosets and liquid crystals, such as low viscosity for ease of processing, good
dimensional stability, high glass-transition temperatures, good thermal stability,
high mechanical properties, the ability to be oriented, and good barrier proper-
ties. Among the applications envisioned for these materials are high performance
resins for composites, optical thin films, and packaging material for microelec-
tronics. The eventual use of LCTs for these types of applications will depend on
the continued investigation and development of their properties.

The first mention of LCTs is in a paper by de Gennes in 1969 (1). The first

experimental investigations occurred in the 1970s (2,3). However, significant num-
bers of publications did not appear in the literature until the 1990s. Early efforts
focussed primarily on the liquid crystalline structure of the monomers or the cured
networks, whereas more recently there have been efforts to examine the evolution
of structure during cure, the nature of the cure process, and a variety of properties.

Molecular Structure of Monomers

Quite a large number of different monomer structures have been synthesized
(4). Reactive end groups utilized include epoxy (5–10), acrylate and methacrylate

Encyclopedia of Polymer Science and Technology. Copyright John Wiley & Sons, Inc. All rights reserved.

140

LIQUID CRYSTALLINE THERMOSETS

Vol. 3

(11–14), maleimide and nadimide (15–21), vinyl (22,23), isocyanate and cyanate
ester (24–27), and ethynyl (28–30). In one case, a dual-curing monomer containing
both acrylate and ethynyl groups has been made (31). The mesogen is generally
aromatic, and may include ester (6–8,10,16,19–21,25–30,32–34), amide (15,18),
or azomethine (35,36) linkages. There may also be a flexible spacer, typically be-
tween the mesogen and the reactive group (10,13,37). Two fairly unusual struc-
tures involve twin mesogen LCTs, in which there are two mesogenic units sepa-
rated by a flexible spacer (34,36,38,39), and LCTs in which the reactive group is
placed between the flexible unit and the mesogen (40). It is also possible to cre-
ate branched monomers that exhibit a liquid crystalline phase (41). While most
monomers exhibit either achiral nematic or smectic phases, by appropriate choice
of the monomer structure it is possible to create chiral smectic (42) and discotic
phases (43). Tables 1 and 2 provide a representative list of monomers that have
been synthesized. Two epoxies that have been extensively compared are shown
below.

Diglycidyloxy-

α-methylstilbene (1) is a liquid crystalline epoxy, while the

diglycidyl ether of bisphenol A (2) is a typical non-liquid-crystalline epoxy.

LCT monomers follow the general rules for liquid crystalline behavior

that have been found for nonreactive low molar mass liquid crystals (44–48).
Thus, bulky substituents tend to destabilize the liquid crystalline phase (16,27–
30,33,35), longer flexible units favor the formation of smectic phases (40,49), and
odd–even effects are seen in the transition temperatures as a function of flexible
spacer length (36,40,49). Rigid mesogens that form liquid crystalline order upon
cure may not be liquid crystalline themselves. For example, 1 is itself a monotropic
nematic, with a melting temperature of 128

◦

C and an isotropic to nematic transi-

tion upon cooling of 95

◦

C (50). Extension of 1 with a flexible unit such as glutaric

acid creates an oligomer which is enantiotropic (8). In this case the structure may
be likened to that of a main chain liquid crystalline polymer containing rigid and
flexible segments. The phase diagram for binary mixtures of LCT monomers may
be calculated using the same techniques as are employed for low molar mass liquid
crystals, although slight discrepancies are found between theoretical and exper-
imental phase diagrams (51). These discrepancies may be accounted for by the
reactive nature of LCTs, whereby a complex mixture of partially reacted species
is formed during the experiment that is not accounted for in the theory. It has
also been found that the type of linkage can have an effect on the phases shown.
For example, Lee and co-workers have examined the effect of an ester group ver-
sus an ether group between the mesogen and the flexible spacer of an epoxy LCT

Vol. 3

LIQUID CRYSTALLINE THERMOSETS

141

Table 1. Representative Epoxy LCTs

Mesogen

End group

References

9

8

8

8

8

8

8

10

10

10

10

35

(R

1

and R

2

may be hydrogen or methyl)

35

52

38

36

(x ranges from 6 to 9)

(52). They found that the ester group leads to a monomer with a smaller nematic
temperature range. As in low molar mass liquid crystals, this can be attributed
to reduced intermolecular interactions due to the electron-withdrawing character
of the ester group, and lower geometrical anisotropy due to the larger size of the
ester group.

142

LIQUID CRYSTALLINE THERMOSETS

Vol. 3

Table 2. Representative Miscellaneous LCTs

Mesogen

End group

References

11

(R is hydrogen or methyl;

n

= 3, 6, 11)

12,14

(R is hydrogen or methyl)

(R is hydrogen or methyl;

n

= 6)

11,13

15

16

(R is hydrogen, methyl, or chlorine)

20

(n

= 5, 6, 8, 9)

22

(R is hydrogen or methyl)

24

(R is methyl or chlorine)

25,26

25,26

25,26

25,26

26

(R is hydrogen)

26

(R is hydrogen)

Vol. 3

LIQUID CRYSTALLINE THERMOSETS

143

Table 2. (Continued)

Mesogen

End group

References

28

(R is hydrogen, methyl, methoxy, or chlorine)

29,30

(R is hydrogen, methyl, methoxy, or chlorine)

90

Cure Behavior

Reaction Kinetics.

The presence of a liquid crystalline phase can have a

dramatic effect on polymerization rates. The activation energy can be lower in the
liquid crystalline phase, implying a higher rate constant (53). For those monomers
which are initially isotropic, isothermal differential scanning calorimetry mea-
surements show an initial exotherm due to the polymerization in the isotropic
phase, followed by a second exotherm due to a rate acceleration when the re-
acting system undergoes a phase transition to a liquid crystalline phase. This
phenomenon appears to be general, as it appears in both chain-growth (54–56)
and step-growth (57–59) systems. Mormann and Br¨ocher found a similar second
exotherm for an epoxy system that was undergoing a smectic to nematic transition
during cure (60). This may be due to the lower viscosity of the nematic phase.

Several studies have compared the kinetic parameters that result from poly-

merization in different phases for chain-growth systems. Douglas and co-workers
used Raman spectroscopy to compare liquid crystalline and non-liquid crystalline
bisacetylene monomers (28). They found that the initial polymerization rate was
higher in the liquid crystalline phase. Other studies have examined the photopoly-
merization of methacrylates and acrylates using differential scanning calorimetry.
Hoyle and co-workers found there was no change in the initial polymerization rate
as a function of phase, but the maximum rate was reached at lower conversions
as the order of the phase increased (56). Guymon and Bowman performed a de-
tailed kinetic study for liquid crystalline and non-liquid-crystalline diacrylates
polymerized in a liquid crystalline solvent (61). Remarkably, they found that the
polymerization rate is higher at lower temperatures, again due to the greater
order present in the lower temperature liquid crystalline phases. A detailed anal-
ysis showed that this rate increase is caused by a decrease in the termination rate
constant rather than by an increase in the propagation rate constant.

Kinetic studies have also been conducted on epoxy systems. For amine cure,

it is generally assumed that the primary amine has a greater reactivity than the
secondary amine. However, two studies found that for liquid crystalline epoxies
the secondary amine is more reactive than the primary amine (62,63). There is no

144

LIQUID CRYSTALLINE THERMOSETS

Vol. 3

complete explanation for this behavior, but it has been suggested that it is caused
by the lower viscosity of the nematic phase. More detailed kinetic analysis has
been done to model the kinetics during the complete reaction. Non-liquid crys-
talline epoxies are typically found to follow an autocatalytic model, in which one
considers the reaction between an epoxide and an amine or anhydride hardener,
with the reaction catalyzed by a species formed during the reaction (eg hydroxyl
group). This model fits the isothermal differential scanning calorimetry data for
both isotropic epoxies, and for liquid crystalline epoxies that show only a single
exothermic peak during cure (57,64). For systems that do show a second peak due
to a phase change during cure, modifications to the autocatalytic model are needed.
The basic concepts inherent in these modifications are that the reaction proceeds
at a different rate in the liquid crystalline phase than in the isotropic phase, and
the isotropic and liquid crystalline phases contribute to the rate according to the
relative proportion of each phase present. Liu and co-workers developed a model
by modifying the autocatalytic model with a rate-enhancement term (57,59) and
also introduced an error function so that the contribution of the rate-enhancement
term increases as the conversion increases. Micco and co-workers also developed
a model by assuming that the reaction rate is autocatalytic in the isotropic phase
but linear with conversion to the liquid crystalline phase (58). They also assumed
a form for the rate of growth of the liquid crystalline phase, and that the isotropic
and liquid crystalline phases contributed to the overall rate in proportion to their
volume fractions. Given the similarities of the two models, it is not surprising that
both are able to model the double exotherm. Unfortunately, these models do not
give much insight into the mechanism behind the rate enhancement, although
Liu and co-workers do propose that their rate-enhancement term is associated
with an increased concentration of reactive groups at the layer boundaries in the
smectic phase.

Network Formation.

Only very little work has been conducted on examin-

ing the effect that the liquid crystalline phase may have on the network formation.
The two studies conducted indicate that in fact the liquid crystalline phase has
no effect on gelation. Jahromi and co-workers examined the viscoelastic behav-
ior of a liquid crystalline epoxy during cure at a single temperature (65). They
found that the critical gelation exponent was equal to 0.5, which is also found
for isotropic epoxies. Cho and Douglas performed measurements on a different
epoxy at multiple temperatures, and found that the conversion at gelation was
independent of cure temperature, as would be expected for a purely statistical
step-growth process (66). Both studies found that the conversion at gelation was
in good agreement with that calculated from standard theories of gelation via step
growth. Thus it appears that, at least in epoxy systems, gelation is independent
of the phase in which the cure reaction occurs.

Phase Evolution during Cure.

Although the liquid crystalline proper-

ties of LCT monomers are easily understood on the basis of low molar mass liquid
crystals, the structures of the resulting networks are much more complicated,
resulting from an interplay among the monomer structure, the effect of cross-
linking, and the cross-linking agent, if one is present. In some cases, there is no
change in the liquid crystalline texture during cure, and the network exhibits
the same liquid crystallinity as does the monomer (28,30,67). Even in such cases,
however, there may be a slight loss of order at the molecular level due to the
cross-linking reaction. Such an effect has been predicted theoretically (68),

Vol. 3

LIQUID CRYSTALLINE THERMOSETS

145

and there is at least some experimental evidence that this may actually occur
(69).

In many cases, however, the liquid crystalline order of the network is not the

same as that of the monomer. Most common is the case in which the order of the
network is actually greater than that of the monomer, in seeming contradiction
with the theoretical calculations mentioned above. The most common example is
1, which is a monotropic nematic liquid crystal as a monomer, and can be cured
from the isotropic phase to give a liquid crystalline network (50,67,70). There are
also other examples of this behavior (38,39,71–74), including systems which are
initially nematic and then cure to give a smectic network (3,67,71). In general
this behavior is explained on the basis of a minimum aspect ratio for the molecule
that is required for liquid crystallinity to occur (28,39). It is assumed that the cure
reaction increases the aspect ratio such that the molecules organize into a liquid
crystalline phase. However, this explanation does not account for transitions
from a monomer that is nematic to a network that is smectic, nor why some
monomers that are initially isotropic form a network that is smectic. Therefore,
it has also been suggested that specific molecular interactions between either
the mesogens or the cross-linking molecules may be operative in the formation of
smectic phases (38,69).

There are also several cases in which the molecular order decreases during

cure, with the formation of a nematic network from a smectic monomer (75,76)
or isotropic networks from both nematic and smectic monomers (28,40,76). This
phenomenon is explained as being due to the disruption of the order due to the
formation of cross-links. For example, Gavrin and Douglas synthesized a series of
bisethynyl LCTs in which the cross-linking group is attached directly to the meso-
gen (40). In this particular case the cross-linking group is not decoupled from the
mesogen through a flexible spacer, and thus if the distance between the molecules
in the liquid crystalline phase is inconsistent with the bonding distance, the liquid
crystallinity will be disrupted. This behavior is therefore similar to the general
requirements for a flexible spacer to decouple the mesogen from the backbone in
side-chain liquid crystalline polymers. However, other bisethynyl LCTs do not lose
liquid crystallinity during cure (28), and thus there are clearly some details of this
behavior that are not yet understood.

Another important factor affecting the network structure, at least in epoxies,

is the cross-linking agent that is used. A study on this factor was conducted by
Mihara and co-workers (77). They examined several combinations of mesogenic
and nonmesogenic epoxies and amines. The most significant result from their
study was that it is possible to create a liquid crystalline network if a mesogenic
amine is used, even if the epoxy used is nonmesogenic. This clearly indicates
the strong role that the hardener may play in determining the liquid crystalline
structure of the network. Another example of this phenomenon occurs with 1
(50,70). When cured with a difunctional amine the resulting network forms a
nematic structure. However, when cured with the unsymmetric tetrafunctional
amine sulfanilamide, the resulting network forms a smectic structure. The authors
propose that this difference is caused by the unequal reactivities of the two amine
groups in sulfanilamide. The aromatic amine tends to react more easily, causing
chain extension with little cross-linking, which is assumed to favor formation of
the smectic phase. Only at the later stages of cure does the sulfonamide group
react, resulting in cross-linking. However, Barclay and co-workers have found

146

LIQUID CRYSTALLINE THERMOSETS

Vol. 3

that 1 and its oligomers cross-linked with the symmetric tetrafunctional amine
methylenedianiline can form either a nematic or smectic phase, depending on the
molecular weight distribution; oligomers with a narrower distribution tend to form
the smectic phase (67). This result suggests that the most important factor may
be the presence of molecular interactions that drive the formation of the smectic
phase, and that a broader molecular weight distribution disrupts the regularity
of that structure, leading to a nematic phase. Similarly, use of an aliphatic acid to
cure 2,6-diglycidyloxynaphthalene results in a nematic network, despite the low
aspect ratio of the epoxy monomer (72). In this case, hydrogen bonding may play
an important role in the formation of the liquid crystalline structure.

It is apparent, then, that the formation of liquid crystalline structure in cured

LCT networks is a complicated process and is currently difficult to predict. Never-
theless, for practical applications it is desirable to at least have knowledge of the
liquid crystallinity that may form as a function of the cure conditions. To achieve
this, a number of workers have created transformation diagrams that show the
liquid crystalline phases that form as a function of cure temperature and cure time
(28,30,36,38–40,70,71,73,78). Figure 1 is one example of these types of diagrams.
One of the interesting questions that arises from these types of diagrams is the con-
version dependence of the phase transformation. Several studies have addressed
this issue by performing independent measurements of conversion versus time at
various cure temperatures, and then converting the time axis to conversion (63,
66,78). They find that the conversion at which the transformation occurs increases
with increasing cure temperature. This is because at higher cure temperatures
the critical molecular length for liquid crystallinity to occur is higher, and thus
a greater conversion is needed. This explanation is supported by experimental
results on other systems, for which an increase in isotropization temperature as
a function of cure time at a single cure temperature has been reported (50).

200

180

160

140

120

100

80

60

40

T

e

mper

ature

, °

C

0

10

20

30

40

50

60

Time, min

K

LC

LC

I

I

Fig. 1.

Transformation diagram for 1 cured with sulfanilamide showing changes in mor-

phology as a function of the cure time. This particular diagram shows the time at which
phase transformations occur when the sample is cured isothermally at a given tempera-
ture. I, isotropic; LC, liquid crystalline; K, crystalline. Reprinted from Ref. 70, Copyright
(1994), with permission from Elsevier Science.

Vol. 3

LIQUID CRYSTALLINE THERMOSETS

147

Properties

Mechanical Properties.

Liquid crystalline polymers are known to have

outstanding mechanical properties due to their microdomain morphology and
their ability to be oriented. It is expected, therefore, that LCTs may have the
same advantages. The effects of orientation are considered in a later section; here
we consider only macroscopically unoriented systems.

As part of their study on orientation effects, Benicewicz and co-workers mea-

sured the tensile modulus of 1(69). They obtained a value of 3 GPa, which is
typical for non-liquid-crystalline epoxies. Earls and co-workers also found flex-
ural modulus values of approximately 3–4 GPa for both 1 and 2(9). Tan and
co-workers found even lower values of modulus for different liquid crystalline
epoxies, although they did not make a direct comparison to isotropic systems (79).
Ortiz and co-workers also found no enhancement in modulus for 1, with the stor-
age modulus in the glassy state actually lower than for an isotropic epoxy and
the compressive modulus the same (80). From this limited data it appears that
the liquid crystalline phase has no advantage for tensile modulus. This may not
be surprising since modulus in the glassy state is governed by bond deformations
and rotations, which will be relatively unaffected by the liquid crystalline state.
Ortiz and co-workers did find, however, an increase in the dynamic modulus in
the rubbery state above what is predicted from rubber elasticity theory (80). This
increase is greater for the smectic phase than for the nematic phase. They ex-
plain this behavior by proposing that deformation due to motion of the cross-links
in an ordered phase will disrupt that ordered phase, resulting in an additional
free energy penalty due to the deformation. They also examined additional com-
pressive properties, and found that the liquid crystalline phase results in a lower
yield stress, no strain-softening region, and a lower strain at failure. They propose
that this behavior occurs because the rigid and extended nature of the network
segments prevents plastic deformation. Similarly, Earls and co-workers found a
slightly lower flexural strength for 1 compared to that for 2(9).

Enhancements in fracture toughness for LCTs are greater than for other

properties. Several studies have examined the fracture behavior of LCT epoxies.
Both Sue and co-workers and Ortiz and co-workers found that 1 cured in the liquid
crystalline phase had a higher fracture toughness than 1 cured in the isotropic
phase (80,81). Robinson and co-workers conducted a detailed investigation, com-
paring 1 and 2 cured with the same curing agent with the same cure cycle (82). By
varying the epoxy-to-hardener ratio, they found that the morphology of 1 changed,
with the domain size decreasing with increasing deviation from balanced stoio-
chiometry. As a result, the fracture toughness for 1 was higher than that for the
isotropic epoxy only at and near the stoichiometric formulation. Carfagna and co-
workers found that this increased fracture toughness extends to fiber-reinforced
composites, although the increase in impact resistance for LCTs was not as great
in the composites as for the neat resin (83). The mechanism for this enhanced
fracture toughness has been investigated by TEM. Sue and co-workers found that
the crack appears to grow preferentially at the boundaries of the liquid crys-
talline domain, resulting in crack deflection and segmenting, with this mechanism
becoming less effective as the domain size decreases (81). Ortiz and co-workers
found that failure occurs by formation of individual microcracks, which grow and

148

LIQUID CRYSTALLINE THERMOSETS

Vol. 3

Fig. 2.

Thin films of (a) 2 and (b) 1, both cured with methylene dianiline, strained in

tension, and observed between crossed polars in an optical microscope. In (a) can be seen
the typical behavior of an isotropic epoxy: a single crack with birefringence caused by a
shear deformation zone ahead of the crack tip. In (b) there are multiple cracks; these cracks
undergo slow, stable propagation due to failure of individual liquid crystalline domains in
front of and at the crack tip. Reprinted with permission from Ref. 84, Copyright (2000),
Kluwer Academic Publishers.

coalesce, eventually resulting in failure (84). This is in contrast to the isotropic
epoxy, which fails by formation and growth of a single crack. Figure 2 shows a
comparison between the two systems. Ortiz and co-workers proposed that the
individual microcracks undergo stable propagation due to failure of individual
liquid crystalline domains (84). These mechanisms are consistent with the results
of Robinson and co-workers (82), in that it would be expected for these toughening
mechanisms to become less effective as the domain size decreases.

Several studies have examined the adhesive properties of LCTs. Ochi and

Takashima, and Carfagna and co-workers both found an increase in the lap shear
strength for an LCT epoxy compared to an isotropic LCT (85,86). However, Frich
and Economy (87) found that the lap shear strength with a titanium substrate
for an LCT that cures by transesterification was lower than that for an isotropic
resin. Given the limited data available, it is not clear whether these differences are
due to the substrates, surface pretreatment, the type of cure reaction, the liquid
crystalline phase, or some other factor. However, most of these studies did find

Vol. 3

LIQUID CRYSTALLINE THERMOSETS

149

that the failure mode for the LCT was cohesive or mixed cohesive and adhesive,
as opposed to only adhesive for the isotropic thermosets. Thus, it appears that
regardless of the lap shear strength, the LCT has a greater bonding strength
to the substrates. In one case both materials failed in a cohesive manner, and
thus the increased lap shear strength appears to be related to the higher fracture
toughness of the liquid crystalline material (86).

Thermal Stability.

Only a few studies have examined the thermal stability

of LCTs. When the monomers are specifically designed to contain no aliphatic
carbons, the thermal stabilities can be very high. For example, Melissaris and co-
workers synthesized rigid rod monomers with ethynyl end groups (88,89), while
Gavrin and co-workers synthesized monomers with phenylethynyl end groups
(90). In both cases degradation temperatures were reported to be at least 400

◦

C

in both air and nitrogen atmospheres. Thermal stability has also been examined
for epoxy LCTs. The thermal stability is affected by the type and concentration of
hardener used; as the aliphatic content of the resin mixture increases, the thermal
stability decreases (91,92). One study showed that an anhydride hardener resulted
in lower thermal stability than an amine hardener, perhaps due to the difference
in cross-link structure (93). In one study liquid crystalline epoxies were compared
directly to 2(94). When the epoxies were cured with an anhydride hardener the
liquid crystalline epoxies showed a higher onset temperature for degradation, but
when an amine hardener was used 2 had a higher onset temperature. In general,
the results suggest that the high thermal stability is a consequence of the types
of bonds present and is not significantly affected by the liquid crystalline phase.

Permeability.

Liquid crystalline polymers are known to exhibit very low

permeability to various permeants due to the packing of molecules in the liquid
crystalline and crystalline states. The low permeability of LCPs can be under-
stood on the basis of a two-phase model, in which it is assumed that the permeant
does not penetrate into the domains, but can only diffuse through the domain
boundaries (95–97). On the basis of these results, it might be expected that LCTs
would also show low permeability. Carfagana and co-workers measured the sorp-
tion isotherm for 1 and found no difference in water sorption between the nematic
and isotropic states (98). In contrast, Earls and co-workers found that 1 cured in
the smectic state absorbed considerably less methylene chloride, methylethyl ke-
tone, dimethyl formamide, and bleach after 30 days compared to that adsorbed by
2; weight gains for 1 were all less than 1%, while weight gains for 2 ranged from
1.0% (bleach) to 26.6% (methylene chloride)(9). Feng also found considerable dif-
ferences between 1 and 2 in the smectic phase (99). Using a permeation technique
to measure the transport of water vapor through thin films, he found that the
permeability, diffusion coefficient, and sorption coefficient were all substantially
lower for 1. From the limited data available, it appears that the liquid crystalline
state present in the cured resin has an important effect on the diffusion proper-
ties, with the smectic phase being considerably less permeable than the nematic
or isotropic phases.

Optical Properties.

All of the optical applications of LCTs have consid-

ered the cholesteric phase. The cholesteric phase is interesting because thin films
of a cholesteric liquid crystal oriented such that the helical axis is perpendicular to
the plane of the film will selectively reflect circularly polarized light with a wave-
length determined by the pitch of the helix. For low molar mass liquid crystals,

150

LIQUID CRYSTALLINE THERMOSETS

Vol. 3

films are typically created by mixing a nematic liquid crystal with small amounts
of a chiral molecule. The pitch of the helix is determined by both the concentration
and the temperature, and thus they are utilized as simple temperature sensors.
The goal of creating a cholesteric LCT is driven by the desire to permanently fix
the helical pitch, and thus the reflected wavelength, for use in optical devices.

There are several approaches that have been used to create cholesteric LCTs.

These include polymerizing nematic LCTs in the presence of a nonpolymerizable
chiral molecule (100–104), polymerizing a non-liquid-crystalline monomer in the
presence of a nonpolymerizable chiral molecule (105), and copolymerizing nematic
and chiral LCTs (100,106). As with low molar mass liquid crystals, in general
the pure cholesteric compounds do not reflect light in the visible range, although
there is at least one example of a diacrylate which reflects red light when it is not
polymerized (106).

Several studies have examined the properties of the cross-linked cholesteric

LCTs and proposed some interesting applications. Hikmet and Zwerver showed
that removal of the nonpolymerizable molecule leads to an irreversible change in
the helical pitch, and thus the wavelength of reflected light (101). This change
could be induced locally by writing with a laser beam, suggesting that these ma-
terials could be used for optical storage. Ishihara and co-workers used cholesteric
LCTs as optical notch filters that filter light of specific wavelengths for displays,
allowing better color purity for those displays (105). Heynderickx and Broer also
considered cholesteric LCTs for display applications (104). In their case they used
a cholesteric LCT film as a compensation foil to eliminate the wavelength depen-
dence of transmission for a supertwisted nematic display.

Orientation of LCTs in External Fields

The liquid crystalline order of LCTs allows them to be oriented by external fields.
Orientation may be induced at a surface, or by electric or magnetic fields. Cur-
ing the material subsequent to, or during, orientation leads to a cured thermoset
which has some degree of macroscopic order. Unlike liquid crystalline polymers,
shear fields and mechanical deformation in the uncured state are not considered
to be as effective, because of the rapid relaxation of the monomers and the diffi-
culty in maintaining such fields during the cure process. Nevertheless, mechanical
deformation has been shown to be effective at orienting lightly cross-linked mate-
rials by heating above their T

g

and subsequently cooling to lock in the orientation,

although the ability to orient decreases dramatically with increasing cross-link
density and the orientation is lost upon heating above T

g

(107,108). The orienta-

tion process tends to enhance certain properties of the cured LCT, although the
potential benefits have not yet been fully explored.

Surface Induced Orientation.

Liquid crystals are well known to orient

at surfaces (109). This effect is utilized in liquid crystal displays, in which orienta-
tion of the liquid crystal at a surface is transmitted throughout a thin film. Often
this orientation can be achieved simply by rubbing a polymer such as polyimide
repeatedly along a given direction. This rubbing process results in orientation
of the liquid crystal along the rubbing direction. Given that this technique works
well for low molar mass liquid crystals, it is not surprising that it has been utilized

Vol. 3

LIQUID CRYSTALLINE THERMOSETS

151

to align thin films of LCTs (12,22,23,56,110). The technique appears to be very
general, and has been applied to epoxies, vinyl ethers, and methacrylates, using
both thermal and photoinitiated polymerization. The orientation in these systems
can be measured using either infrared spectroscopy or X-ray scattering. The re-
sulting orientation is expressed in terms of the second moment of the orientation
function, or the orientation parameter, which is given by

P

2

=

1
2

(3

cos

2

α − 1)

(1)

where

α is the angle a molecule makes with respect to the alignment direction and

the angular brackets denote an average over all molecules. The limiting values
for

<P

2

> are −0.5 for perfect orientation perpendicular to the alignment direc-

tion, 0 for no orientation, and 1.0 for perfect orientation parallel to the alignment
direction. Values for

<P

2

> resulting from surface-induced orientation typically

range from 0.6 to 0.94, indicating a high degree of alignment.

An interesting application of surface-induced alignment of LCTs is orienta-

tion at the surface of fibers in fiber-reinforced composites. Adams and Mallon have
shown that a low molar mass liquid crystal is oriented at the surface of a carbon
fiber parallel to the fiber direction (111), while Sue and co-workers found that 1
can be oriented along the fiber direction, depending on the cure schedule and the
epoxy/hardener formulation (112). These findings raise the possibility of creat-
ing fiber-reinforced composites with controlled matrix orientation and/or tailored
interfaces, which may improve the fiber/matrix interface and result in improved
properties (see R

EINFORCEMENT

).

Electric-Field-Induced Orientation.

Despite the fact that electric field

orientation of liquid crystals is an important technology used, for example, in liq-
uid crystal displays, there has been almost no work on the electric field alignment
of LCTs. This may be because electric fields are effective only for thin films; the
high field strengths required, on the order of 10

4

V/cm or greater, can lead to dielec-

tric breakdown in thick samples. Nevertheless, although bulk orientation cannot
be obtained via electric fields, use of electric fields does provide a complement to
the surface-induced techniques described above.

The only description of electric field orientation comes out of work by Ober

and co-workers (37,113). They used ac electric fields, which allowed them to control
orientation either parallel or perpendicular to the field direction. This is caused
by an electrohydrodynamic effect, by which there exists a critical field frequency,
below which the molecule orients parallel to the electric field and above which the
molecule orients perpendicular to the electric field. Ober and co-workers were able
to show control over the direction of orientation in a cyanate ester LCT by changing
the frequency of the applied field during the cure (113). In contrast, an epoxy LCT
flipped from a parallel to a perpendicular orientation during isothermal cure at
a given applied frequency (37). This was attributed to a change in the critical
frequency due to the increase in viscosity during cure.

Magnetic-Field-Induced Orientation.

Liquid crystals may be oriented

in magnetic fields, even though they are diamagnetic, due to the anisotropy in
their diamagnetic susceptibility. The advantage of magnetic fields over surface-
induced orientation or electric fields is that samples are not confined to thin films.

152

LIQUID CRYSTALLINE THERMOSETS

Vol. 3

In fact, orientation in magnetic fields is more efficient in thicker samples, due to
anchoring effects at walls. The result is that magnetic fields are the only way to
obtain samples of highly cross-linked materials on which bulk properties can be
measured.

de Gennes has described the fundamental physics behind the orientation of

liquid crystals in magnetic fields (109). Essentially, we can consider an isolated
rod placed in a magnetic field. This rod experiences a torque due to the presence
of the field, given by (109,114)

L

M

= −

1
2

χ B

2

sin 2

θ

(2)

where

χ is the anisotropy in the diamagnetic susceptibility parallel and perpen-

dicular to the molecular axis (

χ

par

-

χ

perp

), B is the magnetic field strength, and

θ is the angle of the molecular axis with respect to the magnetic field direction.
Most liquid crystals have a positive

χ, meaning that the molecular axis aligns

parallel to the applied field. Opposing this magnetic torque is a viscous torque,
given by

L

V

= − γ

1

d

θ

dt

(3)

where

γ

1

is the rotational viscosity coefficient. Essentially, this viscous torque is

caused by the drag exerted on the molecule by the surrounding medium. The time
dependence of the orientation process can be obtained by assuming that steady
state has been reached, i.e. the sum of these torques is zero. Solving the resulting
differential equation leads to

tan

θ = tan θ

0

exp



− t

τ



(4)

where

θ

0

is the initial angle of the molecule when the field is first applied, and

τ

is a relaxation time given by

τ =

γ

1

χ B

2

(5)

There are two important assumptions in this approach. The first is that the

viscosity is a constant. Clearly for LCTs this cannot be true, as the viscosity in-
creases during cure. It is straightforward to include a time-dependent viscosity in
the model, although it makes the solution of the differential equation more com-
plicated. The implications of this time-dependent viscosity will be discussed below.
The other assumption is that this is a monodomain sample, with all molecules at
the same angle with respect to the magnetic field. To account for polydomain sam-
ples, it is possible to integrate equation 4 over all possible domain orientations.
When this is done, the distribution function for domain orientations is given by
(114)

ρ(θ,t) =

sin

θ

0

sin

θ

d

θ

0

d

θ

(6)

where the relationship between

θ

0

and

θ is defined by equation 4. Finally, the

model prediction for average orientation angle of the domains,

< cos

2

θ >, is

Vol. 3

LIQUID CRYSTALLINE THERMOSETS

153

cos

2

θ =

π/2

0

ρ(θ,t)sin θ cos

2

θ dθ

π/2

0

ρ(θ,t)sin θ dθ

(7)

which can then be used to calculate

<P

2

> using equation 1.

Experimentally, it has proven quite feasible to orient LCTs in magnetic fields.

Two classes of materials have been studied: epoxies, which are thermally poly-
merized in the presence of a magnetic field (69,79,107,115–118). and acrylates or
methacrylates, which are oriented in a magnetic field and then photopolymerized
(12,56,119,120). Most of the work understanding the factors affecting the degree
of orientation have been conducted on epoxies. From the model described above,
we see that the factors that could improve orientation are higher field strength;
lower initial viscosity and slower viscosity increase; longer time in the magnetic
field; and a more anisotropic diamagnetic susceptibility. However, very few studies
have examined these factors in detail. Benicewicz and co-workers have shown that
orientation increases for 1 cured with sulfanilamide system up to a field strength
of approximately 10 T (69). Lincoln and Douglas examined the combined effects
of field strength, time in the field, and B-staging (prereaction) (118). They found
that orientation increases with higher field strength, longer time in the field, and
less B-staging (due to the lower viscosity).

The liquid crystalline phase present during the orientation process is also an

important factor. Theoretically, the decrease in energy upon orientation of a single
molecule in the gas phase is orders of magnitude less than the thermal energy,
and thus a single molecule is randomized by thermal fluctuations. In a liquid crys-
talline phase, however, motions of the molecules are coupled, and thus the total
energy decrease of the system is much greater than the thermal energy. In addi-
tion, it has been shown theoretically that packing constraints in a smectic liquid
crystal cause orientation to be considerably more difficult than in a nematic (121).
Thus, we would expect that under identical conditions nematic LCTs would show
the greatest degree of orientation, smectic LCTs would show a lower degree of ori-
entation, and isotropic thermosets would show no orientation. These expectations
are confirmed by experimental results. Hoyle and co-workers photopolymerized a
methacrylate LCT at various temperatures in a 0.53-T field (56). Substantial ori-
entation was observed when polymerization was conducted at low temperatures
in the nematic phase. However, no orientation was observed at high temperatures,
when the monomer existed in the isotropic phase. Barclay and co-workers exam-
ined 1 and oligoethers of 1, cured with methylene dianiline (67,107). Compound
1 showed much lower orientation compared to the oligoethers, because 1 itself is
initially isotropic, while the oligoethers are initially nematic. Thus, the oligoethers
have more time for orientation to occur, and presumably also have lower viscosity
since orientation can occur before any chain branching or cross-linking can occur
due to the cure reaction. Jahromi used deuterium NMR to show that the rate of
orientation is lower in the smectic phase than in the nematic phase (115).

Very little work has been done to understand the kinetics of the orienta-

tion process. The model described above has been used to examine the factors
affecting the degree of orientation (122). The rate of orientation is determined

154

LIQUID CRYSTALLINE THERMOSETS

Vol. 3

0

4000

8000

12000

Time, s

0

10

20

30

40

50

60

70

, deg

0

10

20

30

40

50

60

70

12000

8000

4000

0

Time, s

θ

, deg
θ

Fig. 3.

Model calculations of the magnetic field orientation of liquid crystals. Calculations

were conducted using equations 2 and 3. In (a) comparison is made between a nonreactive
liquid crystal with a constant viscosity and a reactive liquid crystal with an exponentially
increasing viscosity due to the cure reaction —— reactive;

unreactive. In (b) is seen

the effects of magnetic field strength for the reactive liquid crystal

0.01 T;

0.025 T; – – - 0.05 T;

0.1 T. From Ref. 122. To convert T to gauss, multiply by 10

4

.

by a competition between the magnetic field and the viscosity. Figure 3a show
the effect of the time-dependent viscosity as predicted by the above model. For a
liquid crystal with a constant viscosity, the viscosity only serves to retard the ori-
entation process; thus at long times the monodomain sample becomes completely
oriented in the direction of the magnetic field. However, when the viscosity in-
creases exponentially with time, there comes a point where the magnetic field
strength is not sufficient to overcome the viscosity, and orientation ceases. In
real systems, this point may correspond to the gel point. Figure 3b shows that if

Vol. 3

LIQUID CRYSTALLINE THERMOSETS

155

1

0.8

0.6

0.4

0.2

0

−0.2

Orientation parameter

2

3

4

5

6

7

8

9

10

Modulus

, GP

a

Fig. 4.

Effect of magnetic field strength on the tensile modulus of compound 1 cured with

sulfanilamide. The line is a quadratic fit to the data, which indicates that the modulus is
determined by the average orientation of the molecules. Reprinted with permission from
Ref. 69, Copyright (1998), American Chemical Society. To convert GPa to psi, multiply by
145,000.

the field strength is high enough complete orientation can be achieved before the
viscosity increase overcomes the magnetic field. Unfortunately, there have been no
detailed comparisons to date between this model and experimental results, and so
the quantitative accuracy of this model cannot be evaluated. However, predicted
trends do correspond to the experimental results described above.

A few studies have examined the effect of orientation on bulk properties.

Figure 4 shows the change in tensile modulus with orientation for 1 cured with
sulfanilamide. Overall, the tensile modulus can be increased three times by orien-
tation. The authors interpret the quadratic dependence of the modulus on orienta-
tion parameter as indicating that the modulus depends on the average orientation
of the smectic layers (69). Tan and co-workers also found an increase in tensile
modulus upon orientation (79). It has also been shown that orientation leads to an
anisotropic coefficient of thermal expansion (low thermal expansion in the direc-
tion of orientation) (69,107), as well as a higher tensile strength and elongation
at break (79). In contrast, experiments using a non-liquid-crystalline epoxy show
no enhancements in properties after curing in a magnetic field (123), further con-
firming that the liquid crystalline phase is needed for orientation to occur.

Future Directions for Research

The synthesis of LCT monomers and understanding of their cure behavior ap-
pears to be fairly well established. Although further effort may be needed to

156

LIQUID CRYSTALLINE THERMOSETS

Vol. 3

identify LCT monomers appropriate for specific situations, the general princi-
ples of how molecular structure affects the phases present, and the development
of different phase structures during cure, is well understood. Future work is likely
to be focussed on properties and applications of LCTs. For example, only relatively
little work has been conducted on bulk mechanical properties or permeability of
solvents or gases in LCTs. Important fundamental questions remain regarding
the mechanism of enhancement in these properties over conventional isotropic
thermosets, as well as the origin of differences in properties between different liq-
uid crystalline phases (eg nematic vs smectic). Considerable work is also needed
to fully understand and control the orientation of LCTs using external fields. Fi-
nally, applications for LCTs are not yet developed. A few optical devices have been
constructed using LCTs, but general application of LCTs to optical applications
has not been demonstrated. Similarly, the use of LCTs in composite structures
remains limited in scope. Substantial effort is needed to show whether LCTs can
exhibit performance benefits that will justify their use in future applications.

BIBLIOGRAPHY

1. P. G. de Gennes, Phys. Lett. A 28, 725 (1969).
2. L. Liebert and L. Strzelecki, C. R. Acad. Sci. Paris 276, 647 (1973).
3. S. B. Clough, A. Blumstein, and E. C. Hsu, Macromolecules 9, 123 (1976).
4. A. Shiota and C. K. Ober, Prog. Polym. Sci. 22, 975 (1997).
5. C. Carfagna, E. Amendola, and M. Giamberini, Prog. Polym. Sci. 22, 1607 (1997).
6. M. Giamberini, E. Amendola, and C. Carfagna, Macromol. Chem. Phys. 198, 3185

(1997).

7. B. A. Rozenberg and L. L. Gureva, in S. M. Aharoni, ed., Synthesis, Characterization,

and Theory of Polymeric Networks and Gels, Plenum Press, New York, 1992, p. 147.

8. M. Giamberini, E. Amendola, and C. Carfagna, Recent Res. Dev. Polym. Sci. 2, 363

(1998).

9. U.S. Pat. 5,270,405 (1993), J. D. Earls, R. E. Hefner, and P. M. Puckett (to The Dow

Chemical Company).

10. S. Jahromi, J. Lub, and G. N. Mol, Polymer 35, 622 (1994).
11. U.S. Pat. 5,078,910 (1992), M. H. Litt (to Edison Polymer Innovation Corp.).
12. C. E. Hoyle, T. Watanabe, E. Brister, and J. W. Whitehead, Polym. Prep. 34, 703 (1993).
13. M. H. Litt, W. T. Whang, K. T. Yen, and X. J. Qian, J. Polym. Sci., Part A: Polym. Chem.

31, 183 (1993).

14. R. A. M. Hikmet, J. Lub, and P. M. Vanderbrink, Macromolecules 25, 4194 (1992).
15. A. E. Hoyt and B. C. Benicewicz, J. Polym. Sci., Part A: Polym. Chem. 28, 3403 (1990).
16. A. E. Hoyt and B. C. Benicewicz, J. Polym. Sci., Part A: Polym. Chem. 28, 3417 (1990).
17. A. E. Hoyt, B. C. Benicewicz, and S. J. Huang, ACS Symp. Ser. 435, 198 (1990).
18. U.S. Pat. 5,198,551 (1993), B. C. Benicewicz and A. E. Hoyt (to The United States of

America as represented by the United States Department of Energy).

19. U.S. Pat. 5,114,612 (1992), B. C. Benicewicz and A. E. Hoyt (to The United States of

America as represented by the United States Department of Energy).

20. K. Kallal-Bartolomeo, J. C. Milano, J. L. Vernet, and B. Gallot, Macromol. Chem. Phys.

201, 2276 (2000).

21. H. R. Kricheldorf and T. Krawinkel, Macromol. Chem. Phys. 199, 783 (1998).
22. R. A. M. Hikmet, J. Lub, and J. A. Higgins, Polymer 34, 1736 (1993).
23. H. Andersson, U. W. Gedde, and A. Hult, Polymer 33, 4014 (1992).

Vol. 3

LIQUID CRYSTALLINE THERMOSETS

157

24. G. G. Barclay, C. K. Ober, K. I. Papathomas, and D. W. Wang, Macromolecules 25,

2947 (1992).

25. W. Mormann and J. Zimmermann, Polym. Mater. Sci. Eng. 66, 498 (1992).
26. W. Mormann, Trends Polym. Sci. 3, 255 (1995).
27. W. Mormann and C. Kuckertz, Macromol. Chem. Phys. 199, 845 (1998).
28. E. P. Douglas, D. A. Langlois, and B. C. Benicewicz, Chem. Mater. 6, 1925 (1994).
29. H. R. Kricheldorf and A. Gerken, High Perform. Polym. 9, 75 (1997).
30. D. A. Langlois, B. C. Benicewicz, and E. P. Douglas, Chem. Mater. 10, 3393 (1998).
31. J. W. Schultz, J. Bhatt, R. P. Chartoff, R. T. Pogue, and J. S. Ullett, J. Polym. Sci., Part

B: Polym. Phys. 37, 1183 (1999).

32. R. A. M. Hikmet, J. Lub, and A. J. W. Tol, Macromolecules 28, 3313 (1995).
33. W. Mormann and C. Irle, Macromol. Symp. 122, 159 (1997).
34. K. Kurschner and P. Strohriegl, Liq. Cryst. 27, 1595 (2000).
35. W. Mormann, M. Broche, and P. Schwarz, Macromol. Chem. Phys. 198, 3615 (1997).
36. E. J. Choi, H. K. Ahn, J. K. Lee, and J. I. Jin, Polymer 41, 7617 (2000).
37. A. Shiota and C. K. Ober, Macromolecules 30, 4278 (1997).
38. A. Shiota and C. K. Ober, Polymer 38, 5857 (1997).
39. A. Shiota and C. K. Ober, J. Polym. Sci., Part A: Polym. Chem. 34, 1291 (1996).
40. A. J. Gavrin and E. P. Douglas, Macromolecules 34, 5876 (2001).
41. W. Mormann and J. G. Zimmermann, Liq. Cryst. 19, 481 (1995).
42. B. C. Baxter and D. L. Gin, Macromolecules 31, 4419 (1998).
43. C. D. FavreNicolin and J. Lub, Macromolecules 29, 6143 (1996).
44. M. J. S. Dewar and R. M. Riddle, J. Am. Chem. Soc. 97, 6658 (1975).
45. M. J. S. Dewar and A. C. Griffin, J. Am. Chem. Soc. 97, 6662 (1975).
46. S. A. Haut, J. P. Schroede, and D. C. Schroede, J. Org. Chem. 37, 1425 (1972).
47. S. L. Arora, J. L. Fergason, and T. R. Taylor, J. Org. Chem. 35, 4055 (1970).
48. M. J. S. Dewar and R. S. Goldberg, J. Org. Chem. 35, 2711 (1970).
49. A. J. Gavrin, PhD dissertation, University of Florida, Gainesville, Fla., 2000.
50. Q. H. Lin, A. F. Yee, H. J. Sue, J. D. Earls, and R. E. Hefner, J. Polym. Sci., Part B:

Polym. Phys. 35, 2363 (1997).

51. A. E. Hoyt and S. J. Huang, J. Macromol. Sci.—Pure Appl. Chem. A 32, 1931 (1995).
52. J. Y. Lee, J. S. Jang, S. M. Hong, S. S. Hwang, and K. U. Kim, Polymer 40, 3197

(1999).

53. G. Schnurpfeil, A. Harder, H. Schroder, D. Wohrle, A. Hartwig, and O. D. Hennemann,

Macromol. Chem. Phys. 202, 180 (2001).

54. C. E. Hoyle, C. P. Chawla, D. Kang, and A. C. Griffin, Macromolecules 26, 758 (1993).
55. C. E. Hoyle and T. Watanabe, Macromolecules 27, 3790 (1994).
56. C. E. Hoyle, T. Watanabe, and J. B. Whitehead, Macromolecules 27, 6581 (1994).
57. J. P. Liu, C. C. Wang, G. A. Campbell, J. D. Earls, and R. D. Priester, J. Polym. Sci.,

Part A: Polym. Chem. 35, 1105 (1997).

58. G. Micco, M. Giamberini, E. Amendola, C. Carfagna, and G. Astarita, Ind. Eng. Chem.

Res. 36, 2976 (1997).

59. J. P. Liu, C. C. Wang, G. A. Campbell, J. D. Earls, and R. D. Priester, J. Polym. Sci.,

Part A: Polym. Chem. 36, 1457 (1998).

60. W. Mormann and M. Brocher, Polymer 40, 193 (1999).
61. C. A. Guymon and C. N. Bowman, Macromolecules 30, 1594 (1997).
62. A. Mititelu, T. Hamaide, C. Novat, J. Dupuy, C. N. Cascaval, B. C. Simionescu, and P.

Navard, Macromol. Chem. Phys. 201, 1209 (2000).

63. E. Amendola, C. Carfagna, M. Giamberini, and G. Pisaniello, Macromol. Chem. Phys.

196, 1577 (1995).

64. P. Punchaipetch, V. Ambrogi, M. Giamberini, W. Brostow, C. Carfagna, and N. A.

D’Souza, Polymer 42, 2067 (2001).

158

LIQUID CRYSTALLINE THERMOSETS

Vol. 3

65. S. Jahromi, W. A. G. Kuipers, B. Norder, and W. J. Mijs, Macromolecules 28, 2201

(1995).

66. S. Cho and E. P. Douglas, Macromolecules 35, 4550 (2002).
67. G. G. Barclay, C. K. Ober, K. I. Papathomas, and D. W. Wang, J. Polym. Sci., Part A:

Polym. Chem. 30, 1831 (1992).

68. D. L. Kurdikar, H. M. J. Boots, and N. A. Peppas, Macromolecules 28, 5632 (1995).
69. B. C. Benicewicz, M. E. Smith, J. D. Earls, R. D. Priester, S. M. Setz, R. S. Duran, and

E. P. Douglas, Macromolecules 31, 4730 (1998).

70. Q. H. Lin, A. F. Yee, J. D. Earls, R. E. Hefner, and H. J. Sue, Polymer 35, 2679 (1994).
71. J. J. Mallon and P. M. Adams, J. Polym. Sci., Part A: Polym. Chem. 31, 2249 (1993).
72. M. Giamberini, E. Amendola, and C. Carfagna, Polym. Eng. Sci. 39, 534 (1999).
73. J. S. Grebowicz, Macromol. Symp. 104, 191 (1996).
74. W. Mormann and M. Brocher, Macromol. Chem. Phys. 197, 1841 (1996).
75. S. Osada, K. Tsunashima, T. Inoue, and S. Yano, Polym. Bull. 35, 505 (1995).
76. S. Osada, S. Yano, K. Tsunashima, and T. Inoue, Polymer 37, 1925 (1996).
77. T. Mihara, Y. Nishimiya, and N. Koide, J. Appl. Polym. Sci. 68, 1979 (1998).
78. S. M. Hong, S. S. Hwang, and J. Y. Lee, J. Polym. Sci., Part B: Polym. Phys. 39, 374

(2001).

79. C. B. Tan, H. Sun, B. M. Fung, and B. P. Grady, Macromolecules 33, 6249 (2000).
80. C. Ortiz, R. Kim, E. Rodighiero, C. K. Ober, and E. J. Kramer, Macromolecules 31,

4074 (1998).

81. H. J. Sue, J. D. Earls, and R. E. Hefner, J. Mater. Sci. 32, 4031 (1997).
82. E. J. Robinson, E. P. Douglas, and J. J. Mecholsky Jr., Polym. Eng. Sci., 42, 269 (2002).
83. C. Carfagna, G. Meo, L. Nicolais, M. Giamberini, A. Priola, and G. Malucelli, Macro-

mol. Chem. Phys. 201, 2639 (2000).

84. C. Ortiz, L. Belenky, C. K. Ober, and E. J. Kramer, J. Mater. Sci. 35, 2079 (2000).
85. M. Ochi and H. Takashima, Polymer 42, 2379 (2001).
86. V. Ambrogi, C. Carfagna, M. Giamberini, E. Amendola, and E. P. Douglas, J. Adhes.

Sci. Technol. 16, 15 (2002).

87. D. Frich and J. Economy, J. Polym. Sci., Part A: Polym. Chem. 35, 1061 (1997).
88. A. P. Melissaris, J. K. Sutter, M. H. Litt, D. A. Scheiman, and M. A. Schuerman,

Macromolecules 28, 860 (1995).

89. A. P. Melissaris and M. H. Litt, Macromolecules 27, 2675 (1994).
90. A. J. Gavrin, C. L. Curts, and E. P. Douglas, J. Polym. Sci., Part A: Polym. Chem. 37,

4184 (1999).

91. C. Farren, M. Akatsuka, Y. Takezawa, and Y. Itoh, Polymer 42, 1507 (2001).
92. M. G. Lu, M. J. Shim, and S. W. Kim, J. Appl. Polym. Sci. 75, 1514 (2000).
93. C. Carfagna, E. Amendola, and M. Giamberini, Macromol. Chem. Phys. 195, 2307

(1994).

94. W. F. A. Su, K. C. Chen, and S. Y. Tseng, J. Appl. Polym. Sci. 78, 446 (2000).
95. V. Vittoria, R. Russo, F. Decandia, P. L. Magagnini, and B. Bresci, J. Polym. Sci., Part

B: Polym. Phys. 29, 1163 (1991).

96. D. H. Weinkauf and D. R. Paul, J. Polym. Sci., Part B: Polym. Phys. 30, 817 (1992).
97. D. H. Weinkauf and D. R. Paul, J. Polym. Sci., Part B: Polym. Phys. 30, 837 (1992).
98. C. Carfagna, E. Amendola, M. Giamberini, G. Mensitieri, M. A. Delnobile, and A. G.

Filippov, Polym. Eng. Sci. 35, 137 (1995).

99. J. Feng, PhD dissertation, University of Florida, Gainesville, Fla., 2001.

100. A. Stohr and P. Strohriegl, J. Macromol. Sci.—Pure Appl. Chem. A 34, 1125 (1997).
101. R. A. M. Hikmet and B. H. Zwerver, Liq. Cryst. 13, 561 (1993).
102. R. A. M. Hikmet and B. H. Zwerver, Liq. Cryst. 12, 319 (1992).
103. R. A. M. Hikmet and B. H. Zwerver, Mol. Cryst. Liq. Cryst. 200, 197 (1991).
104. I. Heynderickx and D. J. Broer, Mol. Cryst. Liq. Cryst. 202, 113 (1991).

Vol. 3

LIQUID CRYSTALLINE THERMOSETS

159

105. S. Ishihara, F. Yokotani, Y. Matsuo, and K. Morimoto, Polymer 29, 2141 (1988).
106. J. Lub, D. J. Broer, R. A. M. Hikmet, and K. G. J. Nierop, Liq. Cryst. 18, 319 (1995).
107. G. G. Barclay, S. G. McNamee, C. K. Ober, K. I. Papathomas, and D. W. Wang, J.

Polym. Sci., Part A: Polym. Chem. 30, 1845 (1992).

108. A. Shiota and C. K. Ober, J. Polym. Sci., Part B: Polym. Phys. 36, 31 (1998).
109. P. G. de Gennes and J. Prost, The Physics of Liquid Crystals, 2nd ed., Clarendon Press,

Oxford, 1993.

110. D. J. Broer, J. Lub, and G. N. Mol, Macromolecules 26, 1244 (1993).
111. P. M. Adams and J. J. Mallon, Mol. Cryst. Liq. Cryst. 208, 65 (1991).
112. H. J. Sue, J. D. Earls, R. E. Hefner, M. I. Villarreal, E. I. Garcia-Meitin, P. C. Yang, C.

M. Cheatham, and C. J. G. Plummer, Polymer 39, 4707 (1998).

113. H. Korner, A. Shiota, T. J. Bunning, and C. K. Ober, Science 272, 252 (1996).
114. J. S. Moore and S. I. Stupp, Macromolecules 20, 282 (1987).
115. S. Jahromi, Macromolecules 27, 2804 (1994).
116. C. K. Ober and G. G. Barclay, Mater. Res. Soc. Symp. Proc. 227, 281 (1991).
117. C. B. Tan and B. M. Fung, J. Polym. Sci., Part B: Polym. Phys. 39, 915 (2001).
118. D. M. Lincoln and E. P. Douglas, Polym. Eng. Sci. 39, 1903 (1999).
119. J. W. Schultz, R. P. Chartoff, and J. S. Ullett, J. Polym. Sci., Part B: Polym. Phys. 36,

1081 (1998).

120. J. W. Schultz and R. P. Chartoff, Polymer 39, 319 (1998).
121. O. Parodi, Solid State Commun. 11, 1503 (1972).
122. E. P. Douglas, Polym. Prep. 40, 498 (1999).
123. R. Gerzeski, Polym. Eng. Sci. 39, 2150 (1999).

E

LLIOT

P. D

OUGLAS

University of Florida

LITERATURE OF POLYMERS.

See I

NFORMATION

R

ETRIEVAL

.

LIVING POLYMERIZATION, ANIONIC.

See A

NIONIC

P

OLYMERIZATION

.

LLDPE.

See E

THYLENE

P

OLYMERS

, LLDPE.

LOW DENSITY POLYETHYLENE.

See E

THYLENE

P

OLYMERS

, LDPE.