Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
PROBLEMS AND ISSUES: RECOM MENDATIONS FOR FURTHER WORK MACROMOLECULAR DESIGN AND SYNTHESIS Sidechain LCPs Although not suited for structural use, sidechain LCPs have potential as functional materials. The anisotropic organization of pendant groups can be intimately related to function, as, for example, in the nonphotobleachable colored film materials derived from cholesteric sidechain polymers (Shannon, 1984~. In this case and in general, mesomorphic structure is readily frozen into a glassy mesophase without crystallization as the polymer is cooled below its glass transition temperature (T~. Retention of mesophase positional alignment and dipolar orientation upon electric field poling while cooling below To can lead to bulk noncentro- symmetry in a mesomorphic system (Meredith et al., 1982) e Development of properties dependent on an acentric structure, such as certain NL0 effects, piezoelectricity, and pyroelectric~ty, is possible if the structural chemistry of the polymer is designed properly. To the degree an anisotropic arrangement of functional species can contribute to enhancement of these effects, a mesogenic polymer is advantageous. Ferroelectric behavior has been observed for smectic C sidechain LCPs (Shibaev et al., 1984~. Electro-optical devices based on ferroelectric LCs are known (Clark and Lagerwall, 1984~. Although usually slower in response than low-molar-mass ferroelectric LCs, advantage can be taken of their polymeric nature. Since surface-stabilized ferroelectric LCs are in a sense "self-poling," a combination of ferroelectric and, for example, NL0 properties in a single material would be attractive. Lightly cross-linked LCPs, to date primarily sidechain LCPs, can be elastomeric above T~ (Finkelmann et al., 1981~. These elastomers exhibit orientation of mesogenic pendant groups upon application of stress to the network. LCP elastomers have been envisioned as optical waveguides and as 71
72 selective barrier materials (membranes). In situ photopolymerization of sidechain LCPs in the mesophase has been recently demonstrated and is perceived (Broer et al., 1988; Hoyle et al., 1988) as a processing option that may lead to high-speed formation of LCP films and coatings. In addition, masking techniques used with in stun photopolymerization are seen as potentially useful for integrated optics, displays, and optical information discs. Laser-addressed smectic sidechain LCPs have been described for information- storage applications (Hopwood and Coles, 1985 ~ . Historically, the synthesis of sidechain LCPs has been primarily free- radica1 vinyl polymerization of acrylate or methacrylate monomers . Recently, combined sidechain and mainchain LCPs have been described (Reck and Ringsdorf, 1985 ~ based on a polyester mainchain. There has been rapid development of synthetic methodology toward introduction of new mainchains for sidechain LCPs including polysulfones (Braun et al ., 1987 ), polyesters (Griffin et al ., 19 8 8 ) , and polyurethanes (Tanaka and Nakaya, 19 8 8 ~ Polysiloxanes (Finkelmann and Rehage, 1980) and, more recently, polyphosphazenes (Singler et al., 1987) have been used for reactions in which a mesogenic pendant group is grafted onto the polymer backbone. Sidechain LCPs from these preformed polymers and from new polymer backbones offer interesting auxiliary properties, such as variability in T~, solubility, stability, etc. Hence, it is recommended that attention be paid to novel functional sidechain (and mainchain) LCPs. Opportunities for exploiting the interplay of function and chemical structure in these anisotropic systems should be taken where their structural anisotropy and glassy mesophase formation can lead to enhanced performance. Thermotrop~c LCPs Among the serious structural issues arising in LCP polyester syntheses are the elucidation and control of the mer sequence along the polymer backbones (Economy et al., 1989; Muhlebach et al., 1988~. At best, one can only see dyads via high-resolution NMR in the solid state or in solution (where soluble). These tend to confirm X-ray studies, which conclude that the sequences are essentially random (in other than the trivial simple A-A B-B case) (Blackwell and Biswas , 1986 ~ . Given the heterogeneous nature of polymerizations involving free terephthalic acid as well as differences in condensation rates, blocky sequences would be anticipated in the absence of extensive transesterification. Clearly, the latter is an important accompanying reaction as confirmed recently (Economy et al. , 1989 ; Jin 1989~ ; controlled sequence distributions were shown to randomize rapidly by interesterification on heating. To control the sequence distribution, new polymerization techniques and/or catalysts and transesterification inhibitors will have to be developed, as well as improved methods to measure the resulting distr~butions. Until these are accomplished, the structure-property relationships associated with backbone sequence and subsequent design of possibly improved polymers will be severely hampered.
73 Monomer Costs The reason for the commercial and military importance of LCPs resides in their unique properties. However, their commercial future and availability is intimately tied to lowering their cost. In turn, this is largely determined by the cost of the monomers required for their manufacture. The LCPs that are most significant from a commercial and/or military viewpoint are derived essentially from all-aromatic monomers. The significant monomers for commercial and near-commercial thermotropic polyesters include A-B as well as A-A and B-B types. The latter include the dials, such as 4,4'-biphenol (in Amoco's Xydar~ and S~mitomo's Ekonol@), hydroquinone (in ICI's SRP) and phenylhydroquinone (Du Pont), and the diacids, including terephthalic and isophthalic acids. The significant A-B monomers are 4-hydroxybenzoic (in most LCPs) and 6-hydroxy-2-naphthoic acids (in Hoechst-Celanese's Spectral. Of the above, the lowest-cost monomers are the diacids, at substantially less than $1.00/lb. Next in price are hydroquinone (potentially) and p-hydroxybenzoic acid, at less than $2.00/lb. The most expensive of the thermotropic LOP monomers are 4,4'-biphenol and 6-hydroxy-2-naphthoic acid, at somewhere between $3.00/lb and $10.00/lb. Phenylhydroquinone is probably included in the latter range. Of the lyotropic LOP monomers, terephthalic acid and its acid chloride (equivalent) are under $.00/lb, p-phenylenediamine is in the $2.00/lb range, the meta, pare hybrid diamines are probably in the $2.00 to $5.00/lb range, and far and away the most expensive of the monomers, by at least an order of magnitude, are the multifunctional monomers required for the heterocyclic PBX LCPs. Future Cost Challenges The cost challenges inherent in developing future thermotropic LCPs will be either (a) to develop higher-temperature properties utilizing the lower- cost hydroquinone or (b) to develop low- cost syntheses for the higher- temperature monomers or (c) to design and synthesize new low-cost aromatic monomers. An example of a very significant breakthrough via the second approach would be a direct coupling of phenol to produce 4,4'-biphenol in one step rather than via the existing processes, which either sulfonates the biphenyl followed by alkali fusion or couples 2,6-di-t-butylphenol at the 4 position, to form the 3,3' ,5,5' -tetra-t-butyl-4,4' -biphenol, and subsequent dealkylation. It should be noted that direct coupling of phenol has been achieved biologically but with no selectivity, all available carbon positions coupling at random. An example of the third approach would be a similar one- step coupling of benzoic acid to yield the 4,4'-bibenzoic acid. As for the lyotropics, the biggest challenge is the development of lower- cost routes to the benzobisthiazole and benzobisoxazole polymers and by lower- cost routes to their multifunctional monomer precursors. Because of the very high price and therefore limited availability of these monomers, any advances in synthetic approaches would have a marked beneficial effect on their availability and on the future of the extremely high-perfo~mance materials
74 that can be made from them. Of considerable significance is the recent announcement by Dow that it is preparing a benzobisoxazole polymer in pilot- plant quantities. UNDERSTANDING AND THEORY Semiflexible LCPs Liquid crystal formation by semiflexible polymers has been modeled by extensions of the Onsager method alluded to in Chapter 2. The relevant controlling parameter revealed by these theories (Doi and Edwards, 1986; Odijk, 1986) is the ratio of the persistence length to diameter of the polyme chain. Little experimental verification of these theories is yet in the literature. Dramatic predictions concerning the influence of macromolecular flexibility on the order parameter and concentration regime of biphasic stability are contained in these newer theories. These predictions should be tested with carefully obtained data on polymer systems that conform to the assumptions of the developments . No totally general models exist yet for the anticipation of liquid crystallinity of semiflexible polymers at all concentrations. In particular, the thermotropic behavior is only poorly understood. The primary difficulty in developing a comprehensive picture of the thermodynamic aspects of LCPs lies in the fact that a high degree of coupling exists between all degrees of freedom in the system, internal and external to the polymer chains. Contrary to the situation in polymer melts and in amorphous systems, where chains can assume all conformations and the external constraints on a chain are essentially isotropic, in the liquid crystalline phase the necessity for the macromolecule to conform to the anisotropic spatial requirements of its environment reduces drastically its freedom in conformation (internal) and orientation (external). Chains in crystalline materials are similarly restricted, of course, but there the reduction of fully developed degrees of freedom for the chain is so drastic that in most cases only one conformation and one packing arrangement can exist; as a consequence, the analysis of crystalline systems has been mastered decades ago, and very simple rules can be applied (e.g., Natta's "equivalence principle," which states that in a polymer crystal the chain conformation must be a repeated sequence of local conformations). In liquid crystalline phases the molecules in a mesomorphic phase must adopt orientations, configurations, and a "packing'' arrangement compatible with the weak, but not insignificant, constraints of symmetry and density of the mesophase. In consequence, both intra- and inter-molecular degrees of freedom, static and dynamic, are highly coupled, between molecules as well as within individual chains. This coupling is effective not only in thermotropic systems but also in lyo tropic ones, as long as the concentration of polymer exceeds a critical value.
75 PolvdispersitY and Blends of Polymers If systems with more than one component are involved in phase equilibria, one can, in general, expect an unequal distribution of the different species between the phases; LCPs are no exception. Three interesting situations exist: polymers not uniform with respect to molecular weight, mixtures of flexible and rigid chains, and LCP blends. To date, the first two situations have been addressed with some success: · As prepared most LCPs are not uniform with respect to molecular weight and therefore are multicomponent systems; they may fractionate, especially in solution. The higher-average-molecular-weight species are found in the anisotropic phase; the lower-average-molecular-weight species accumulate in the isotropic phase. Computations have been performed for a variety of molecular weight distributions (see Flory, 1984, and citations therein). · Mixtures of ideal random coils and rigid rods also fractionate strongly (Flory, 1984), even when the chains are composed of "identical" segments and are of identical molecular weight. The rigid species strongly prefer the anisotropic phase. No theoretical treatment seems to have addressed the question of the blending of different mesogenic polymers. If both types of chains are constitutionally uniform, of equal rigidity and sufficiently long, the problem, to zeroth approximation, can be viewed in analogy to the blending of random coiling chains; in both cases the configurational entropy of mixing is very small and the enthalpy of mixing determines if blending is possible. One would therefore expect that LCPs mix only rarely. However, in flexible chains, miscibility can often be achieved by blending copolymers, and there is no reason why similar effects cannot be exploited in mesogenic polymers. The situation in LCPs is complicated, however, by the fact that the different species probably would be of different rigidity (exhibit different persistence lengths); one might speculate that for this reason blending would become even rarer than in flexible polymer mixtures, but only initial attempts at theoretically illuminating the situation has been put forward to date (DeMeusse and Jaffe, 19881. Thermodynamic treatments of polymer-polymer mixing and of polymer mesophase formation have been treated by Flory (1986) and others (Flory and Ronca, 1979; Doi and Edwards, 1986; Odijk, 19861. The application of these concepts to LCP-containing blends is just beginning to appear in the literature. There is no reason to expect that these concepts will not provide an effective framework for understanding the phenomena observed in LCP-based polymer mixing. Comparison of LCP behavior in blending with that of low- molecular-weight LCs should also prove instructive. It is to be expected that miscibility between LCPs will be rare, as it is with conventional polymers. In contrast, most low-molecular-weight LCs of a given type are miscible with each other. The in-depth understanding of the physical chemistry of LCP blending will be an important cornerstone in defining the ultimate utility of LCP blends and should be rigorously pursued.
76 Dynamics The dynamics of rigid rod systems have been investigated with intensity (Chandrasekhar, 1977; Doi and Edwards, 1986), and quantitative results are available for moderately concentrated solutions (below the transition to the nematic state). In the bulk, similar problems exist as for flexible chains. To date, only single polymer systems with chains of uniform molecular weight and flexibility (usually stiff rods) have been addressed. The following is known: · The rotational constant in dilute (isotropic) solution of rod-like polymers grows with chain length approximately like Dr ox ln~x) /X3 . Delis is a very rapidly decreasing function of the aspect ratio x ; long rigid rods rotate very slowly. · In semidilute solutions (above the concentration where frequent intermolecular contacts occur, but still in the isotropic regime) of rigid- rod polymers, the rotational diffusion is attenuated by an additional dependence of the approximate form x-2 with respect to the one in dilute solution. · In concentrated (i.e., anisotropic) solutions, and in the bulk phase, no rotational diffusion effectively exists. · The translational diffusion coefficient parallel to the rod axis is roughly twice that perpendicular to that axis in dilute (isotropic) solution. Both diffusion constants grow as D ~ ln~x)/x. This indicates a much less rapid decline than in the case of the rotational diffusivity. · For semidilute (isotropic) solutions of rigid-rod polymers, the translational diffusion in the direction of the rod is roughly that for the same rod in dilute solution, but translation perpendicular to the rod is negligible. · Very little is known about translational diffusion in the anisotropic solutions and in the bulk phase. Many other dynamic phenomena have been addressed; of particular interest are the viscosity of solutions as a function of concentration (which follows the experimentally observed cusp curve) (Hermans, 1967) and the degree of order as a function of the concentration and external fields in anisotropic phases. Little quantitative information is available, however, in the latter cases. Rheolo~v The anisotropic orientation in melt-processed molded parts cannot be predicted because a complete continuum theory for LOP rheology is not available. The classical theory of nematic liquid rheology developed by Leslie (1966, 1968) and Ericksen (1960, 1961) is inadequate for polymers because of the simplifying assumptions in the derivation: The stress is
77 assumed to be linear in the instantaneous rate of deformation, and the entropic free energy associated with local director gradients, which leads to an elastic "Frank" stress, is assumed to be quadratic. The theory does not take chain flexibility into account, and magnetic field experiments on thermotropic nematic melts (Moore and Denn, 1988) have indicated that there is creep in the Frank stress. The theory thus stands in relation to LCPs (particularly lyotropic solutions of very rigid molecules) in much the same way as the Newtonian fluid to conventional flexible polymers. Much can be gained by use of the Leslie-Ericksen theory, but results cannot be expected to be quantitative and may even be incorrect in qualitative terms for flows well outside the applicable range of the theory. One important qualitative feature of LOP flow that does seem to be contained in the Leslie-Ericksen theory because of the presence of the entropic Frank elastic stress is the tendency of the director sometimes to adopt an orientation that is transverse to the plane of shear (Rey and Denn, 1989; Beris and Edwards, 1990~. A continuum theory of LCPs by Doi (1981) is incomplete in that it does not contain any elastic contribution to the stress resulting from local director gradients; the Doi theory thus apparently cannot predict any phenomena associated with rapid spatial changes in orientation, although an extension of the theory that includes director gradient, has been reported (Beris and Edwards, 1990~. Neither the Leslie-Ericksen nor the Doi theory has been applied to the flow of fluids with domain-like textures; it is unlikely that the Doi theory in its present form can be applied here because of the limitations already cited. The theological feature of LCPs that has been most exploited in applications other than fiber formation is the very high degree of shear thinning at all deformation rates (see Kulichikhin, 1989, and Muir and Porter, 1989, for recent reviews). It is this property that makes thermotropic melts attractive for molding applications with complex shapes and small passages. Qualitative models (Marrucci, 1984; Wissbrun, 1985) based on the relative motion of domains predict shear thinning at low deformation rates, but no adequate theory exists here either. It is possible that the presence of microcrystals in the melt is a factor in the shape of the flow curve, as well as long transients that are observed under conditions of apparent thermal stability. The self- organizing feature of LCPs is an asset or a liability , depending on the shaping flow of interest. Predictive ability is lacking in any event, because the rheology is not understood in a fundamental way. Simulations of flow and orientation distributions in complex parts are unlikely to be correct in the absence of an adequate theological theory, except in the case of a uniformly extensional flow of sufficient strength to effect uniaxial orientation (in which case no theory is needed). Progress in melt fabrication of molded parts that exploit the orientability of LCPs is unlikely until sufficient understanding of LOP behavior has been achieved to allow the modeling of geometrical and filling schemes that will lead to orientation distributions that approximate macroscopic isotropy and contain no rapid changes in orientation; the latter is particularly serious because of the poor self-adhesion of rigid polymers. It is not obvious that these problems can be overcome by conventional melt processing.
78 CONCLUSIONS · Rigid rod theories are in reasonably good order; macromolecular flexibility is manageable. ~ There are deficiencies in theories regarding blends of LCPs and LCPs with flexible polymers, and consequently there are no guidelines for experimental work in this area. · A lack of understanding of chain dynamics and rheology of LCPs is preventing solutions to processing problems such as the weld line, adhesion, and orientation development in molds. RECOMMENDATIONS ~ Theoretical modeling leading to predictable distinctions between conventional isotropic phases and LC phases based on local primary and secondary chain structure is necessary to understand and develop new LCPs; intensified research should be encouraged. · More interactive research among theoreticians and experimentalists on aspects of blending LCPs is needed. ~ Theoretical studies of the fundamental dynamics and the rheology of LCPs should be encouraged, with accompanying experimentation on the relation between rheology and microstructure for the development and optimization of processing methods. The implications of available theories of LOP rheology should continue to be explored, but processing conclusions need to be examined with care because of the incompleteness of the existing theories. PROCESSING Lvotropic LCPs Lyotropic LCPs have been the subject of intensive investigations. Molecular composites- a molecular dispersion of (lyotropic) rigid rod polymers in a (conventional) polymer matrix have dominated the question of processing rigid rod LCPs. This novel state of matter remains elusive, however. The intricacies of retaining isolated rod-like molecules from dilute solution into the solid state are unknown, unpublished, proprietary, or a combination of these. Coagulants are critical to forming the molecular composite; their effect on both polymers needs to be addressed before fabricating and processing can be controlled. Another problem lurking in the background concerns the retention of the level of molecular dispersion during use of such composites.
79 Aside from the intensive effort to fabricate molecular composites, there has been one more significant area of activity in processing lyotropic LCPs (apart from fiber production). This concerns attempts to fabricate biaxial structures- films from PBX polymers. While there are reports of success at Foster-Miller using proprietary schemes, this work has been limited for the most part to bench-top scales (R. Lusignea, private communication). A major difficulty with evaluating the potential of this technology appears to be the severely limited availability of the PBX polymers. Pilot plant scale-up will be necessary before a meaningful assessment of film production is possible. However, it would appear that a viable interim solution could be explored- namely, investigating the feasibility of fabricating biaxial structures with commercially available lyotropic LCPs such as Ke~rlar~ aramid. (See also the discussion of film formation in Chapter 2.) Thermotropic LCPs There are fundamental difficulties in traditional processing techniques (injection molding, extrusion, rotational molding, etc.~. This may be generally attributed to intrinsic problems associated with highly ordered polymer melts. Self-adhesion in polymer melts, which is an important factor in the strength of parts that contain weld lines, is governed by diffusion of chains across interfaces. The rigid molecules of LCPs must diffuse a much greater distance to achieve adhesion than flexib] e polymers. It is for this reason that self - adhesion in LCPs is poor, both at weld lines and in regions of rapid orientation change. The "weld line" problem may be insurmountable at the molecular level for highly anisotropic melts and will require an engineering solution based on mold design. Herein we focus on novel processing possibilities for thermotropic LCPs. Solid-State Forming In studies of the extrusion of powder preforms of Xydar~ (a terpolymer of p-hydroxybenzoic acid, terephthalic acid, and biphenol), workers at Alcoa found (Zaidi, 1988) that defect-free rods could be cold-extruded from the preforms only if the preform was annealed properly, the die was designed with the correct angle, and the temperature and rate of extrusion were held in the range that gives metal-like strain-hardening flow characteristics. This suggests that the eons titutive behavior of solid LCPs as a function of temperature and pressure should be a critical area of research. It is also clear that the influence of pressure on solid-solid transitions would be an important conjunctive area of research between chemical structure and novel processing. The explanation for cold-forming-induced optical clarity is still an issue. Contributions resulting from the small size of the crystalline domains in the formed material and the elimination of shrinkage voids have been mentioned. In the case of the LCPs, there is some indication that cold-formed shapes may be free of the skin-core effect and may enjoy outstanding machinability and wear characteristics relative to melt-formed material (Zaidi, 1988).
80 From a general point of view, solid-state forming does have its drawbacks. It often involves heavy, unconventional equipment and may require very expensive tooling. The parts may exhibit poorer dimensional stability than melt-formed ones, and extra allowance is often required for ''spring-back" of the parts after removal from the dies. The severe orientation introduced by some forming processes may result in the tendency of the product to split already a difficulty with LCPs. Because LCPs are relatively expensive polymers, the savings in energy and cycle time associated with solid-state forming may not have a significant effect on the overall cost of the part. The use of powder preforms, which appears to be a way around the splitting problem, may hold little promise for exploitation of the ultimate strength properties of LCPs. If this proves to be the case, the entire combination" LCP fabricated with solid-state extrusion -becomes decidedly unattractive. Clearly, to accrue the full advantages of the preform extrusion technology, the constitutive behavior of solid LCPs as a function of temperature and pressure must be understood. In addition, the influence of pressure on solid-solid transitions would be an important conjunctive area of research for developing the connections between chemical structure and novel processing (Hsiao et al., 1988~. CONCLUSIONS LYotropics An important objective of LCP processing technology is to develop strong materials for ultralight primary structures that are difficult to detect by radar and can withstand elevated temperatures. In the case of composites, the temperature resistance and tensile properties of fibers must also be combined with outstanding compressive properties and adhesion in fabrication schemes. For the LCP fibers to be effective in high-performance structural composites, they should also exhibit compressive strength that approaches the compressive properties of current carbon fibers. Uniaxial structures prepared from liquid crystalline precursors (melts or solutions) exhibit outstanding tensile properties. For some time it was believed that the liquid crystalline state of the precursor during processing (spinning) was essential for achieving the molecular alignments and the perfection of structure that yield almost theoretical modulus and exceptional strengths properties far above those of fibers produced from isotropic melts or solutions (Economy et al., 1970; Aharoni and Sibilia, 1978~. Later, it was shown that the isotropic solutions of semirigid polymers produced at a low rate of elongational flow in the spin-way yield relatively weak fibers as spun. At high rates of deformation, however, the semirigid polymers yield properties close to those of aramids. More recently, it was shown that even very flexible polymers, such as PE and PVA, can be converted into exceptionally strong fibers without becoming obviously liquid crystalline during the process of converting the isotropic solution into fiber. The quiescent liquid crystalline state is, therefore, beneficial for the preparation of strong fibers without after drawing but it
81 is not essential for all polymers. These observations must be borne in mind in assessing processing schemes and/or potential polymers for novel applications. Thermotropics Solid-state forming of LCPs is an immature area. It can be anticipated that many of the advantages of solid-state forming found with conventional semicrystalline polymers will accrue for LCPs. It is clear that there are unanticipated, unique properties that can be obtained by solid-state forming of powder preforms, but the basic science of this process as well as those using melt-formed preforms (e.g., sheets) needs to be investigated. The practical outcome of funding in this area would be an increase in the availability of high-performance LOP products for use in defense-related engineering projects. Not to be overlooked is the "knowledge" product: The exploration of solid-state forming techniques could lead to information concerning important phenomena such as the poor compressive strength of highly oriented LCPs. RECOMMENDATIONS ~ Fabrication of biaxial film using readily available lyotropic LCPs should be investigated. · Other routes (nonmesomorphic gels) to high-modulus fibers should be explored. · Mold-design studies and experiments aimed at the weld line problem should be encouraged. · Solid-state forming of LCPs needs to be explored sufficiently to establish patterns allowing the performance of these novel processes to be compared with the extensive body of information already available for semicrystalline, random-coil polymers, as well as the definition of any advantages unique to LCPs. ~ For electro-optical applications, electric poling and associated fabrication processes need further study. MECHANICAL PROPERTIES Tensile Properties Remote from major transitions, axial properties of solid-state filaments formed from LCPs are excellent and approach the theoretical limit in tensile modulus and exhibit extraordinary high tensile strength. Some typical properties of commercial fibers were shown in Figure 3.1. Transverse
82 properties, however, are poor, often 1 to 2 orders of magnitude lower than axial properties. This is inherent to highly anisotropic materials with a hierarchical morphology (Figure 2.13), and lateral weakness is expected to occur theoretically. These poor compressive properties (see, for example, Figure 4.1) limit the utility of LCPs in materials applications. 5.0 ce CD - I 1.0 G Oh LLI > ~ 0.5 co G o C: 0.1 0.1 O ocfromsin9lefiberte~s c from composite data / /6 0.5 1.0 GRAPHITE ABPBO l NTP ~ 1 ! O / 1 1 ~ i'\. '1 ~ =0.3 G(r=0.98) 1 / ~C / O 5.0 10 TORSION MODULUS (G Pa) FIGURE 4.1 Compressive strength versus torsion modulus for some rigid rod polymers. The outstanding problem is this: Can compressive properties of LCPs be improved without sacrificing desirable tensile properties? A major question in this regard is whether the fault is at the molecular or morphological level. Contemporary thinking suggests problems at a macroscopic level, i.e., interfibrillar bonding in the hierarchical fiber structures. This idea is reinforced because cross-linking (strengthening lateral intermolecular interactions) has not relieved the problem of poor compressive strength. Conversely, it has been argued by Jaffe and Yoon (1987) that essentially all of the mechanical properties of highly oriented LCPs can be explained on the basis of molecular parameters and that morphological features play only a minor role. This is probably not true in complex structures such as moldings and biaxial films. The issue remains to be resolved, however.
83 Cross-linking and blending of LCPs with random coil polymers (molecular composites) have been considered the most promising approaches to enhancing the transverse properties and compressive strength of LCP fibers. Cross- linking has been studied extensively with random coil fibers, such as nylons and polykethylene terephthalate), but the results fall short of expectations because the small improvements (if any) in compressive properties were offset by substantial losses in tensile properties. No successful composite technology based on cross-linked fibers and no successful cross-linking of LCP fibers has been reported as yet. A recent study (Grubb and Kearney, 1990) of the effect of formaldehyde cross-linking on properties of gel-drawn EVA offers a good overview of problems associated with this type of modification of uniaxial structures. A large and probably inevitable decrease in tensile modulus on cross-linking is particularly noteworthy. Unfortunately, the authors did not include the data on strength that usually show a much larger decrease on cross-linking than modulus. In principle, the solution of the compressive strength problem through development of molecular composites appears to be very attractive. It must be noted, however, that authors invariably stress the improvements with LCP- random coil blends, but do not address the problem of compressive strength. Since a successful solution of this problem would be promptly revealed in technical or patent literature, it can be assumed that, heretofore, blending of LCPs and random coil polymers has failed to produce significant improvements in compressive strength. Consequently, we can speculate that poor adhesion between phases and the relatively coarse dispersions achieved so far have prevented successful development of a LCP-based system with a high compressive strength. Thermal Behavior The temperature fall-off of properties exhibited in Figure 4.2 is often characteristic of solids derived from thermotropic LCPs (see also Table 3.3~. A substantial decrease in the tensile modulus is encountered between room temperature and temperatures characteristic of the a-transition in a particular LCP. This behavior may retard the utilization of LCP-based materials in cases where the LCP narrowly fulfills modulus criteria. Improvements in this area often lead to thermally intractable polymers, in turn fail processing requirements. Weld Line Injection molding of three-dimensional items runs into the 'weld line problem.'" Differently oriented director fields in LCPs do not adhere well to themselves. Moreover, when neighboring macromolecular axes are not parallel, interdiffusion is retarded. This fabrication problem, which is severe with thermotropic LCPs, requires clever mold design to ensure intimate mixing of director fields.
84 Tensile data HBA/HNA 100 ce C, C) o of 80 60 40 20 it_ O I I t I I -100 -50 0 50 U 73/27 -- A 73/27 ~A 30/70 i; 100 150 TEMPERATURE ~C) FIGURE 4.2 Temperature dependence of dynamic mechanical tensile data (10 Hz) for HBA-HNA copolymers (Davies and Wood, 1988~. CONCLUSIONS ~ Compressive failure currently limits widespread use of LCPs. Failure mechanisms and consequently potential solutions are not within reach. ~ Rapid deterioration of properties of LCPs with increasing temperature is another limitation that restricts use of these materials. LCPs. . RECOMMENDATIONS Conventional molding techniques exaggerate the weld line problem in · A significant research effort should focus on the compressive failure problem. · Experimental and theoretical work on mold designs should be initiated. ~ Thermal properties of LCPs require molecular design solutions, and associated exploratory syntheses should be undertaken.
85 BLENDS AND COMPOSITES Combining LCPs with other materials to Improve overall property balance, exploit unique LCP performance, mitigate property deficiencies, and/or lower cost is an active and fruitful area of applied LCP research that may be divided into four subareas: Fillers for injection-molding resins Continuous-fiber-reinforced composites LCP-containing polymer blends Molecular composites Fillers The major application of the commercial thermotrop~c copolyesters is as a high-performance injection-molding resin. Fillers are employed in these systems to reduce overall part anisotropy, both in terms of flow direction to cross-flow property balance and control of skin-core effects. Fillers are also useful in minimizing part warpage and controlling the coefficient of thermal expansion. The fillers currently employed in LCP moldings are the same as those used with other thermoplastic engineering resins. Optimization has been limited to available materials, and little systematic research into optimizing filled LCPs can be found in the literature. Investigations into the effects of filler surface characteristics and geometry on molded part performance may prove a key to defining the limits to molded LCP part utility and should be encouraged. Fiber-Reinforced Composites The high-modulus, high-strength properties of LCP fibers have made them attractive candidates as reinforcing fibers for composites. Both the~motropic and lyotropic LCP fibers have been evaluated as reinforcing fibers, but published data for systems other than Kevlar~ fiber are sparse. Typical data sets for LCP fibers may be found in Chapter 3. Tables 4.1A and 4.1B show properties of Kevlar~ and PBZT relative to graphite fibers in epoxy composites. The utility of LCP fibers in composites is limited by poor compressive properties and poor matrix adhesion. The physical structure responsible for low compressive strength does, however, provide advantages in damage tolerance and crash-worthiness. For example, para-aramid composites are metal-like in ductile energy absorption and respond to compressive stress noncatastrophically, unlike carbon composites, which have a brittle or catastrophic response (International Encyclopedia of Composites, 1989~. The industry has recognized the benefits of combining the inherent damage tolerance of the LCP fibers with the compressive properties of carbon fiber in the form of hybrid composite structures.
86 TABLE 4.1A Properties of Unidirectional Epoxy* Composite with 60% Fiber Loading, 0 ° Direction . Kev~ar ~Thorny 300 Property 149 Aramid 49 Aramid Graphite Tensile Strength MPa 1450 1500 1420 Kpsi 210 218 206 Tensile Modulus GPa 107 79 133 Mpsi 15.6 11.5 19.3 % Conversion 99 99 97 Strain to Failure % 1.33 1.71 0.90 Compressive Strength MPa 193 234 - Kpsi 28 34 Compressive Modulus GPa 73 66 - Mpsi 10.6 9.5 - Flexura] Strength MPa 634 655 1192 Kpsi 92 95 173 F]exura] Modulus GPa 79 67 116 Mpsi 11.5 9.7 16.8 Interlaminar Shear Strength MPa 57 59 70 Kpsi B.3 B.6 10.2 Thermal Expansion Coefficient 10-6 m/m °C +0.37 -2.33 - *Epon. 828 epoxy cured with NMA/BD~, laboratory impregnated. Source: Riewald, P. G., A. K. Dhingra, and T. S. Chew-. 1987. ICCM and ECCM, Sixth International and Second European Conference on Composite Materials, July 20-24, 1987. F. L. Matthews, N. C. R. Buskell, J. M. Hod~kinson, Bud J. Morton, eds. 5:5362-5370. New York: I:lsevier Applied Science. Du Pant Company Publication. June 1987. Kevlar. 149, Properties arid Characteristics of Yarn and Composites. E-95612.
87 TABLE 4 .1B Properties of Unidirectional Epoxy Compos ite with 60% Fiber Loading* Property H.T. PBZT** H.S. Graphi te Kev] arms Tensile Strength - MPa 1900 2100 1380 - Kps i 270 300 200 Modulus - GPa 190 150 75.8 - Mpsi 28 21 I] Strain % 1.D ~.5 ~.~ Poisson's Ratio 0.4 0.27 0.34 In-pJane Shear Strength - MPa 35 69 - Kpsi 5 lo In-pJane Shear Modulus - GPa I.4 4. ~ - Mpsi 0.2 0.7 - Compressive Strength - MPa 200 1400 275 - Kps i 30 200 40 Modulus - GPa 190 125 75.8 - Mpsi 27 IS I] Strain % 0.12 1.1 0.4 Flexural Strength - MPa 410 1600 - - Kpsi 60 240 - Modu~us - GPa 140 120 - Mpsi 20 18 - Short-Beam Shear - MPa 31 Il0 - Kpsi 4.5 16 - *Hercules 3501-6 epoxy, laboratory impregnated. **Heat treated PBZT. Source: · By, W. C. and J. F. Malone. 1988. Canadian Textile J., pp. 54-63 (April). · Wolfe,- J. F. 1988. Polybenzothiazoles and Polybenzoxazoles. Pp. 601-635, in Encyclopedia of Polymer Science and Engineering, Vol. 11, 2nd Ed. lI. F. Mark, N. M. Bikales, C. G. Overberger, G. Menses, and J. I. Kroschwitz, eds. New York: John Wiley and Sons.
88 The thermotropic copolyesters have been included in some recent studies evaluating thermoplastic resins for potential as thermoplastic matrices for continuous fiber reinforcement. Attractive features of the LCPs are low viscosity, good thermal stability, and high tensile properties. Negative features are poor adhesion to reinforcing fibers, property anisotropy, and poor compressive strength. Of these, the observation that the LCP matrix tends to orient along the direction of the reinforcing fibers has been most significant in dampening enthusiasm. One intriguing concept is the use of a LCP fiber to reinforce itself, almost as a macromolecular composite. Further expansion of LCP fibers into composites is dependent on improving adhesion, a successful approach to the compressive strength problem, and demonstrated cost-effectiveness. Utility of the thermotropic LCPs as matrices is most dependent on the development of effective methods of orientation control. LCP-Containine Polvmer Blends The motivation for considering LCP-containing blends is discussed in Chapter 2. As of 1988, the open literature dealing with LCP-containing polymer blends was sparse and difficult to search. Although it is beyond the scope of this report to examine and review in detail, the patent literature is probably the richest source of information. Useful reviews, however, are starting to appear (Browstow, 1988~. Most of the LCP blend literature is highly observational in nature, and the data, because of the heavy impact of specific processing regimes and the use of difficult-to-obtain polymers, will be difficult to reproduce. Little attention has been given, for example, to separating the effects of polymer degradation or modification (transester- ification, etc.) from the structures inherent in the blend. With the dearth of LCPs available worldwide, the number of LCPs blended with other polymers is necessarily small. Blends of LCP and Conventional Polvmers Most of the work to date and the area with the greatest potential for commercial exploitation involves the blending of LCPs with conventional polymers. Although a few studies of solution blending with Kevlar~ do exist (Tekayanagi et al., 1980), most of the work has centered on melt-blending thermotropic copolyesters (Vectra@, Xydar@) with engineering thermoplastics (PET, PET, etc.~. For convenience, this work may be separated into three blend regions based on LCP content, as shown in Table 4.2. The potential utility of LCPs as a processing aid for high-viscosity conventional polymers was rigorously pursued by ICI in the early 1980s (Cogswell et al., 1983~. Although the desired viscosity lowering appears to be dominated by the ratio of the viscosities of the components of the mixture, LCPs are unique in possessing both high molecular weight and low viscosity. Two modes of behavior have been observed: (a) blends with viscosity that
89 TABLE 4 . 2 LCP - Content Blending Regimes Weight % LCP Genera] Description Key References 0-15 Processing aid, viscosity Cogswell et a]., 1983. reducti on 15-85 Self-reinforcing resins Kiss, 1987; Weiss et al., 1987. In situ composites 85-100 Modified LCP Patents follows "rule of mixtures" based on the components, and (b) blends with viscosity lower than either component. This latter behavior is not understood. The use of LCPs as processing aids should be the easiest blend application to exploit commercially and, ultimately, may serve to render very difficult-to-process thermoplastics useful in common processes. The most alluring blend regime to most researchers is the "in situ" composite where the LCP phase orients during processing to reinforce the plastic part. The effectiveness of this process is a function of the orientation imparted to the LCP in the chosen process. Published micrographs document morphologies ranging from spheres to fibrils. No quantification of morphology or correlations with process conditions have been published. Adhesion between the LCP and conventional polymer phases is clearly poor. During processing, stress transfer appears to be through the tortuosity of the phases, but this important factor has not been evaluated in depth. One consistent result in all studies is that the presence of an LCP phase renders the blend brittle (Kiss, 1987~. This is probably a consequence of the poor interphase adhesion and requires clarification. Blending is unlikely to solve the weld line problem (Kiss, 1987~. Mechanical properties of the blends, especially tensile modulus, follow expectations consistent with simple composite concepts in the absence of adhesion between matrix and reinforcement. For these blends to be useful commercially, the issues of adhesion and morphology control must be resolved. In addition, LCP in this application is in direct competition with glass and other reinforcing fibers, and hence the cost-effectiveness of the LCP approach must be established. Improved process equipment lifetime, reduced weight, and improved processibility are the likely LCP advantages. Unexamined in the literature is the degree to which nonstructural properties of LCP, i.e., very low gas permeability and high solvent resistance, carry over into blends.
go It has been found that the presence of LCP may accelerate and presumably direct the crystallization of conventional polymers (Joseph et al., 1983~. Blending biphasic polymers, such as the PET-polyHBA copolymers, may promote miscibility between the conventional phase of the Aphasic polymer and another miscible conventional polymer, such as X7-G with poly~butylene terephthalate). These phenomena may offer direction in the search for useful compatibilizing agents for LCP-conventional polymer systems. There are examples given in the literature (Froix, 1984) where the presence of small quantities of LCP blended with a conventional polymer results in mechanical properties significantly better than either component. No explanation for this observation has been advanced. The analogy of "introducing the lignin into a woody morphology consisting only of fibrin" is appealing but is not consistent with emerging models of LCP structure or LCP- conventional polymer interactions. As sparse as the data set describing mainchain nematic LCP blends with conventional polymers is, it is rich compared to the almost nonexistent data on the blending of other types of LCPs sidechain polymers, flexible spacer polymers, smectics, etc. LCP-LCP Blends Researchers have studied blends of thermotropic copolyesters with other thermotropic copolyesters, examining both blends of different copolymer ratios with identical chemistry (HBA-HNA type) and blends of different chemistries. Recognizing that each copolymer may be viewed as a blend (chain-to-chain variations, mer sequence variations), it was hypothesized that blending offered the means to "engineer the distribution. " Initial results indicate that this is the case. Through such blending the behavior of both the mesophase and the solid state can be systematically modified. For example, transition temperatures can be shifted and power law indices of viscosity can be changed. These results also strongly imply that the "sequence matching" model of the thermotropic copolyesters in the solid state is more likely correct than the other models suggested in the literature (DeMeusse et al., 1988~. Evidence was accumulated that LCP-LCP miscibility is not universal and that, at least in the mesophase, the basic concepts are consistent with the observed physical chemistry. The importance of this work is that it offers a direction for achieving sufficient insight into the nature of LCP structure- property relationships to design more appropriate molecules for given end uses. Related work indicates that transesterification may be responsible for some of these observations (J. Economy, private communication). Careful evaluation of the data cannot rule out transesterification effects, but it strongly suggests that transesterification is not causal in the observed behavior. As in the case of LCP-conventional polymer blending, little information exists on the blending of LCPs of different inherent chain architecture or mesophase symmetry. Recent publications show phase separation in blends of sidechain nematics with other similar polymers or small-molecule analogues.
91 It is now established that, in contrast to the behavior of low-molecular- weight LCs, LCPs are often immiscible Investigations into the blending of LCPs with other polymers, conventional or LC, are in their infancy, and little is understood In detail. The literature is sparse, highly observational in nature, and very difficult to reproduce. Much of what is published is more an indication of what might be than an accurate data base from which to draw conclusions. It is clear, however , that, until the aches ion of IMP to other polymers and the rules controlling blend morphology are understood, the field will remain highly empirical and unlikely to yield many commercial successes . Conversely, for all the reasons that polymer blending is an attractive route to modified polymer products, blends containing LCPs are especially attractive. Systematic research to understand the physical chemistry of LCP-containing blends is likely to produce results of both commercial and scientific impact. This is true for LCP-conventional blends and all -LCP blend systems . Although initial work should focus on the commercial nematic polymers, other symmetries and phases should not be ignored. The area of compatibilizers ("molecular glue") is another concept worthy of support because of the high payback pa tential . Molecular Composites To date it is unclear whether a true molecular compos ite utilizing a molecularly dispersed LCP has been demonstrated, although materials possessing small agglomerations of rods (diameter of structure less than 50 A' have been produced with very high tensile properties. It is not yet established, however, if these materials offer an advantage in tension. The concepts underlying molecular composite physics are consistent with the concepts of miscible blends; the materials being produced, even at the very small sizes of rod structures observed, fit the definitions associated with immiscible LCP- conventional polymer blends. Molecular composites can be treated with the already established framework of polymer-polymer mixing and do not require new concepts for accurate description. Success of molecular composites will be strongly linked to the economics of the processes and materials employed. Research in this area should be supported until a meaningful evaluation of cost-performance can be performed. CONCLUSIONS The combination of LCPs with other materials to control the balance of properties and improve cost-effectiveness is clearly an important technology area for increasing the overall utility of LCPs . The problems inhibiting the rapid development of this technology are the same as those slowing LCP acceptance in other areas , namely : High property anisotropy in finished parts, Poor compressive strength, Poor adhesion to conventional and mesogenic materials, and High cost.
92 This is not in contradiction to the observations that fillers and blending can mitigate many of these effects; rather, it is an important hint into what the underlying controlling parameters of the science must be. RECOMMENDATIONS From both the scientific and technological points of view, studies of the behavior of LCPs in multicomponent systems should be strongly supported. The following are specifically recommended: · Support research efforts to understand and exploit the combination of processing and fillers needed to provide a range of properties from fully isotropic to anisotropic. · Establish research efforts to understand and improve LOP adhesion to common composite matrices, fillers, and reinforcements, and other conventional and mesogenic polymers. · Establish systematic research efforts to understand and exploit the physical chemistry of LCP-containing polymer blends. NONLINEAR OPTICAL PROPERTIES In spite of the obvious advantages of organic NLO materials, there are major obstacles for utilizing polymers. In the case of second-order NLO applications, some obstacles are these: · The uncertainty with respect to thermal stability of electric field- induced ordering in harsh environments, i.e., greater than 80°C. · The small published data base on NLO properties, physical properties, and processing conditions for making active and passive structures in polymer films, and the sparse literature on second-order NLO properties of LCPs. · The competitive advantage relative to LiNbO3. . The maturity of LiNbO3-based technology. · A lack of polymers designed specifically for SHG as well as device concepts and structures-for utilizing them. Although considerable recent progress has been made in achieving high nonlinear coefficients, additional work ranging from fundamental science to new materials development remains to be done if materials in this category are to reach their technological potential.
93 Liquid crystallinity has been shown to lead to enhanced alignment under certain poling conditions. The utilization of extremely high poling fields on amorphous systems, where a substantial fraction of dielectric saturation can be achieved, makes this less of a potential advantage. The anisotropic optical properties of LCPs would also have to be made extremely uniform to avoid scattering losses in bulk or wavegu~de devices. The main benefit of liquid crystallinity may be one of imparting thermal stability to systems once they are poled. More esoteric applications might also exploit LCs. For example, the ferroelectric smectic C* phase, which is inherently noncentrosymmetric, might exhibit high degrees of macroscopic alignment and the desirable processing characteristics of a polymer. A polymeric system exhibiting this phase may require poling or other treatments to exhibit alignment, but the phase is inherently bistable as opposed to metastable in the case of poled thermoplastic. The smectic C* state combined with chromophore substituents designed for specific NLO applications would be an extremely attractive materials option. For third-order NLO materials, the obstacles for utilization are more formidable: · The state of understanding of third-order NLO properties based on the electronic structure of the constituent molecular or polymeric species is at a more rudimentary level, and a considerable diversity of opinion exists among scientists regarding the fundamental origin of nonlinear responses in these materials. Highly delocalized electronic structures are difficult to design into stable, easily processible polymeric structures. point. . Approaches to increasing the nonresonant X(3) are unclear at this · The stringent linear optical properties, including optical uniformity, transparency, and low scattering losses, combined with large X(3) required for device applications , have not yet been demonstrated, and formidable obstacles associated with -C-H vibrational overtones in the 1.3-pm and 1.55-pm regions remain to be addressed. Practical polymer processing conditions for achieving high degrees of orientation to enhance xt3) in formats suitable for waveguide formation have not yet been demonstrated. Organic materials, because of their large nonresonant X(3) values, may constitute an enabling technology if progress can be made in the areas listed above. Low-dimensional electronic structures tend to exhibit the large values of X(3)' since all of the oscillator strength is confined to one predominant direction in these materials. At the macroscopic level, uniaxial orientation of a material with low-dimensional molecular constituents or chains can increase X(3) by up to a factor of 5. Because of trade-off between n2 and a, it will be important to achieve the factor of 5 by process ing and fabrication
94 methods. Liquid crystallinity and other self-organizational influences and factors may offer routes to alignment, optical uniformity, and stability. CONCLUSIONS ~ Inorganic crystals and thin films have advantages for some NLO applications, by virtue either of their function or the maturity of their technological position, but some may be displaced by organic polymeric thin- film materials because of cost and performance advantages; other applications may be made possible by the unique properties of organic and polymeric materials, assuming that fundamental advances can be made. ~ For second-order NLO processes, poled LCPs containing chromophores with a large molecular hyperpolarizability may offer advantages over amorphous polymers with respect to the degree and stability of the induced second-order nonlinear coefficient at the expense of more complex linear optical proper", es. · For third-order NLO processes, LCPs offer ~ route to achieving uniaxial alignment through processing and the associated potential for a five- fold increase in nonlinear coefficient relative to an electronically identical amorphous system, although considerable progress in fundamental understanding of microscopic processes, materials design and synthesis, and processing will be required to take advantage of this. RECOMMENDATIONS ~ Research initiatives on organic and polymeric materials for second- order NLO addressing the issues of stability and magnitude of alignment by electric field poling, spontaneous or self-alignment, physical property studies, processing, and exploratory device utility should be supported. Priority should be given to interdisciplinary program where molecular and polymer design and synthesis, NLO studies, and polymer physical characterization can be combined. · Research initiatives aimed at improving the microscopic understanding of X(3) and testing models through structure property investigations should be supported. Because of the rudimentary state of theory, the data base of X(3) measurements on organic and polymeric structures should be expanded. Synthesis and characterization of new polymeric systems with an emphasis on increasing X(3) and achieving excellent linear optical properties including transparency at 0.85 ~m, 1.3 am, and 1.55 Am should be supported. Processing studies aimed at the unique requirements of wave guide nonlinear optics should be supported. Interdisciplinary studies of a fundamental nature as well as novel exploratory approaches should be encouraged.
95 REFERENCES Aharoni, S. M. and J. P. Sibilia. 1978. Crystalline transitions and the solid state extrusion of polymers. ACS Polym. Preprints 19~1) :350-354. Beris, A. N. and B. J. Edwards. 1990. Poisson bracket formulation of incompressible-flow equations in continuum-mechanics. J. Rheology 34~1~:55-78. Blackwell, J. and A. Biswas. 1986. Stiff chain persistence lengths in liquid-crystalline copolyesters. Makromol. Chem.-Macromol. Symp. 2:21- 31. Boehm, R. E., D. E. Martire, and N. V. Madhusudana. 1986. A statistical thermodynamic theory of thermotropic linear main-chain polymeric liquid- crystals. Macromolecules 19~9) :2329-2341. Braun, D., R. P. Herr, and N. Arnold. 1987. Liquid-crystalline polysulfones .1. Copolymers of I-alkenes with sulfur-dioxide. Makromol. Chem.-Rapid Commun. 8~7~:359-363. Broer, D. J., H. Finkelmann, and K. Kondo. 1988. Insitu photopolymerization of an oriented liquid-crystalline acrylate. Makromol. Chem.-Macromol. Chem. Phys. 189~1~:18S-194. Browstow, W. 1988. Kunststoffe (German Plastics) 78:411. Clark, N. A. and S. T. Lagerwall. 1984. Ferroelectrics 59:25. Cogswell, F. N., E. P. Griffin and J. B. Rose. 1983. U.S. Patent no. 4,386,174 (ICI). DeMeusse, M. T. and M. Jaffe. 1988. Mol. Cryst. Nonlin. Opt. 157:535. Davies, G. D. and I. M. Ward. 1988. P. 377 in High Modulus Polymers, A. Zacharides and R. Porter, eds. New York: Marcel Dekker, Inc. Doi, M. 1981. Molecular dynamics and rheological properties of concentrated solutions of rodlike polymers in isotropic and liquid crystalline phases. J. Polym. Sci. 19~2~:229-243. Doi M., and S. F. Edwards. 1986. The Theory of Polymer Dyn~mics. Oxford: Clarendon Press. Economy, J. 1989. Aromatic polyesters of p-hydroxybenzoic acid. Mol. Cryst. Liq. Cryst. 169:1-22.
96 Economy, J ., R. D . Johnson, A. Muhlebach, and J. Lyerla. 1989. ACS Polym. Preprints 2:505. Economy, J., B. E. Nowak, and S. G. Cottis. 1970a. Tractable, high temperature p-oxybenzoyl polymer. ACS Polym. Preprints 11~1~:332-333. Economy, J., B. E. Nowak, and S. G. Cottis. 1970b. Ekonol. A high temperature aromatic polyester. SAMPE J. 6459:21-27. Ericksen, J. L. 1960. Arch. Ration. Mech. Anal. 4:231. Ericksen, J. L. 1961. Trans. Sac. Rheol. 5:23. Field, N. D., R. Baldwin, R. Layton, P. Prayer and F. Scardiglia. 1988. Polymorphism in a liquid- crystalline polyester based on 4,4'-biphenol, terephthalic acid, and p-hydroxybenzoic acid (1-1-2~. Macromolecules 21(7):2155-2160. Finkelmann, H., H. J. Kock, and G. Rehage. 1981. Investigations on liquid crystalline polysiloxanes .3. Liquid-crystalline elastomers A new type of liquid-crystalline material. Makromol. Chem.-Rapid Commun. 2~4~:317 322. See also: Bualek, S. and R. Zentel. 1988. Crosslinkable liquid-crystalline combined main-chain side-group polymers with low glass-transition temperatures. Makromol. Chem.-Macromol. Chem. Phys. 189~4~:791-796 and references therein. Finkelmann, H. and G. Rehage. 1980. Makromol. Chem.-Rapid Commun. 1:31. Flood, J. E. and J. F. Fellers . 1986. SPE ANTEC Proceedings 32:728. Flory, P. J. 1984. Adv. Polym. Sci. 59:1. Flory, P. J. 1956. Proc. Royal Soc. London A234:73. Flory, P. J. and G. Ronca. 1979. Theory of systems of rodlike particles .2. Thermotropic systems with orientation-dependent interactions. Mol. Cryst. Liq. Cryst. 54~3-4~:311-330. Froix, M. 1984. U.S. Patent no. 4,460,735. Griffin, A. C., A. M. Bhatti, and R. S. L. Hung. 1988. Side-chain polymalonate liquid- crystals for nonlinear optics . Mol . Cryst . Liq . Cryst. 155: 129-139. Grubb, D. T. and F R. Kearney. 1990. Modification of gel-drawn poly~vinyl alcohol) fibers with formaldehyde. J. Appl. Polym. Sci. 39~3) :695-705. Hermans Jr. , J . 1967 . Ordered Fluids and Liquid Cryst. 63 : 282 .
97 Hopwood, A. I. and Coles, H. J. 1985. Liquid crystalline polymers as additives to enhance the device properties of low molecular mass liquid crystals. Polym. 26~9~:1312-1318. Hoyle, C. E., C. P. Chawla, and A. C. Griffin. 1988. Photopolymerization of a liquid-crystalline monomer. Mol. Cryst. Liq. Cryst. 157:639-650. Hsiao , B. S ., M. T. Shaw, and E. T. Samulski . 1988 . Pressure- induced phases in a thermotropic polyester. Macromolecules 21~2) :543-545. International Encyclopedia of Composites, Vol. 1. 1989. New York: VCH Publishers, Inc . . Jin, J . I . 1989 . Polymer Preprints 30:481. Joseph, E. G ., G . L. Wilkes and D. G . Baird. 1983 . Thermal and structural studies of flexible and semi-rigid polymeric blends. ACS Polym. Preprints 24 ~ 2~: 304- 305 . Kiss, G. 1987. In situ composites: Blends of isotropic polymers and thermotropic liquid crystalline polymers. Polym. Engr. & Sci. 27~6) :410- 423. Kulichikhin, V. G. 1989. Mol. Cryst. Liq. Cryst. 169:51. Leslie, F. M. 1968. Arch. Ration. Mech. Anal. 28:265. Leslie, F. M. 1966. Quart. J. Mech. Appl. Math. 19:357. Marrucci, G. 1984. P. 441 in Advances in Rheology, Vol. 1, Proceedings of 9th International Congress on Rheology, B. Mena, A. Garcia-Rej on, C. Fangle-Naffale, eds. Cuidad Universitaria, Mexico City. Matheson, R. R. and P. J. Flory. 1981. Statistical thermodynamics of mixtures of semirigid macromolecules -chains with Godlike sequences at fixed locations. Macromolecules 14~4~:954-960. See also: Ronca, G. and D. Y. Yoon. 1982. Theory of nematic systems of semiflexible polymers .1. High molecular-weight limit. J. Chem. Phys. 76 ( 6 ): 3295 - 3299 . Vasilenko, S. V., A. R. Khokhlov, and V. P. Shibaev. 1984. lrheory of liquid-crystalline ordering in melts of macromolecules with stiff and flexible fragments in the main chain .2. Effect of external orientational fields. Macromolecules 17~11~:2275-2279. Meredith, G. R., J. G. VanDusen, and D. J. Williams. 1982. Optical and non-linear optical characterization of molecularly doped thermotropic liquid-crystalline polymers. Macromolecules 15~5~:1385-1389. Moore, R. G. and M. M. Denn. 1988. P. 169 in High Modulus Polymers: Approach to Design and Development, A. Zachariades and R. Porter, eds. New York: Marcel Dekker.
98 Muhlebach , A ., R. D . Johnson , J . Lyerla , and J . Economy . 19 88 . Diad sequence distribution in copolyesters of 4-hydroxybenzoic acid and 6-hydroxy-2-naphthoic acid. Macromolecules 21~10) :3115-3117. Muir, M. C. and R. S. Porter. 1989. Processing rheology of liquid- crystalline polymers - A review. Mol . Cryst. Liq . Cryst. 169: 83 - 95 . Odijk, T. 1986. Theory of lyotropic polymer liquid-crystals. Macromolecules 19 ~ 9 ): 2314 - 2329 . See also: Khokhlov, A. R. and A. N. Semenov. 1981. Physica A108: 645. Khokhlov, A. R. and A. N. Semenov. 1982. Liquld-crystalline ordering in the solution of partially flexible macromolecules. Physica A112~3~:605 614. Khokhlov, A. R and A. N. Semenov. 1985. On the theory of liquid crystalline ordering of polymer-chains with limited flexibility. J. Stat. Phys. 38~1-21:161-182. Onsager, L. 1949. Ann. N.Y. Acad. Sci. 51:627. Reck, B. and H. Ringsdorf. 1985. Combined liquid-crystalline polymers mesogens in the main chain and as s ide groups . Makromol . Chem.-Rapid Commun. 6~4~:291-299. Rey, A. D. and M. M. Denn. 1989. Analysis of transient periodic textures in nematic polymers. Liq. Cryst. 4~4~:409-422. Shannon, P. J. 1984. Photopolymerization in cholesteric mesophases. Macromolecules 17 ~ 9 ): 1873 -1876 . Shaw, M. T. 1980. Cold forming of polymeric materials. Ann. Re~r. Mater. Sci . 10: 19-42 . Shibaev, V. P., M. V. Kozlovsky, L. A. Beresnev, L. M. Blinov, and N. A. Plate. 1984. Thermotropic liquid-crystalline polymers .16. Chiral smectics C with spontaneous polarization. Polym. Bull. (Berlin) 12~4~:299-301. See also: Blinov, L. M., V. A. Baikalov, M. I. Barnik, L. A. Beresnev, E. P. Pozhidayev, and S. V. Yablonsky. 1987. Experimental-techniques for the investigation of ferroelectric liquid-crystals. Liq. Cryst. 2~2~:121 130. Decobert, G., F. Soyer, and J. C. Dubois. 1985. Chiral liquid crystalline side-chain polymers. Polym. Bull. 14~21:179-186. Decobert, G., J. C. Dubois, S. Esselin, and C. Noel. 1986. Some novel smectic Cstar li quid-crystalline side-chain polymers. Liq O Cryst. 1 (4~: 307 - 317 . Zentel, R., G. Reckert, and B. Reck. 1987. New liquid-crystalline polymers with chiral phases. Liq. Cryst. 241~:83-89.
99 Singler, R. E., R. A. Willingham, R. W. Lenz, A. Furukawa, and H. Finkelmann. 1987. Liquid-crystalline side-chain phosphazenes. Macromolecules 20~7~:1727-1728. See also: Kim, C. and H. R. Allcock. 1987. A liquid-crystalline polytorganophosphazene). Macromolecules 20~7~:1726-1727. Tanaka, M. and T. Nakaya. 1988. Liquid-crystalline polyurethanes .4. Polyurethanes containing an azobenzene group in the side-chain. Makromol. Chem.-Macromol. Chem. Phys. 189~4~:771-776. Tekayanagi, M., T. Ogata, M. Morikawa, and T. Kai. 1980. J. Macromol. Sci., Phys. B17: 591. Weiss, R. A., W. Huh and L. Nicolais. 1987. Polym. Engr. Sci. 27:684. Wissbrun, K. F. 1985. A model for domain flow of liquid-crystal polymers. J. Chem. Sac., Faraday Disc. (799:161-173. Wolfe, J. F. 1988. Polybenzothiazoles and Polybenzoxazoles. Pp. 601-635 in Encyclopedia of Polymer Science and Engineering, Vol. 2, 2nd Ed., H. F. Mark, N. M. Bikales, C. G. Overberger, G. Menges, and J. I. Kroschwitz, eds. New York: John W~ley and Sons. Zaidi, M. A. 1988. Application of Metal-Working Techniques to Process Liquid Crystalline Polymers. Report to the Committee on Liquid Crystalline Polymers . Cambridge, Mass . (July 16 ~ .