The Physics of Liquid Crystals SECOND EDITION P. G. pp GENNES and J. PROST Ecole Supérieure de Physique et de Chimie Industrielles de la Ville de Paris CLARENDON PRESS - OXFORD 1993THE INTERNATIONAL SERIES OF MONOGRAPHS ON PHYSICS GENERAL EDITORS J. BIRMAN S. F. EDWARDS C. H. LLEWELLYN SMITH M. REESOxford University Press, Walton Street, Oxford OX2 6DP Oxford New York Toronto Delhi Bombay Calcutta Madras Karachi Kuala Lumpur Singapore Hong Kong Tokyo Nairobi Dar es Salaam Cape Town Melbourne Auckland Madrid and associated companies in Berlin Ibadan Oxford is a trade mark of Oxford University Press Published in the United States by Oxford University Press, New York © P. G. de Gennes and J. Prost, 1993 Alll rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press. Within the UK, exceptions are allowed in respect of any ‘fair dealing for the purpose of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, or in the case of reprographic reproduction in accordance with the terms of licences issued by the Copyright Licensing Agency. Enquiries concerning reproduction outside those terms and in other countries should be sent to the Rights Department, Oxford University Press, at the address above. This book is sold subject to the condition that it shall not, by way of trade or otherwise, be lent, re-sold, hired out, or otherwise circulated without the publisher's prior consent in any form of binding or cover other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Gennes, Pierre Gilles de. The physics of liquid crystals | P. G. de Gennes and J. Prost. — 2nd ed. p. cm. — (The International series of monographs on physics; 83) Includes bibliographical references. 1. Liquid crystals. I. Prost, Jacques. II. Title. III. Series: International series of monographs on physics (Oxford, England); 83. QD923.G46 1993 530.4'29-—de20 93-9154 ISBN 0 19 8520247 Typeset by Integral Typesetting, Great Yarmouth, Norfolk Printed in Great Britain by Biddles Ltd, Guildford and King’s LynnPREFACE TO THE SECOND EDITION The first version of this book came out about twenty years ago. Since then, our knowledge of liquid crystals has expanded considerably: the ‘blue phases’ of cholesterics have been essentially understood; columnar systems have been constructed; nematic polymers of various types have come up; microphases obtained with block copolymers have been clarified. The first book was a hasty introduction to a rapidly growing field. Some friends convinced us that it was worthwhile to transform it into a more detailed review of a mature field. The process of going from a sketch to a full painting is long and hard: it was energetically undertaken by one of us—Jacques Prost—with very little help from the original designer. The final painting is much larger in size: but it remains focused on a few central characters—with special emphasis on smectics and columnar phases. On the one hand, what was formally a single chapter on smectics has been split into four—retaining a spirit similar to what we originally had for nematics. On the other hand, there are many little corners in the painting where we do not shed much new light: surface phenomena, instabilities, electric field effects in nematics, chemical inventions, applications. ... One beautiful figure is not really shown: this is the modern classification of defects, with its remarkable predictions for systems such as biaxial nematics. The basic reasons for these omissions are (a) the need to keep to a ‘human’ size, (b) the recent appearance of many books where these features are duly emphasized. As in most large classical paintings, this book is really a collective work, where the influence of famous ‘schools’ can be clearly recognized—and where some pieces may turn out to be done by some future Veronese. We acknowledge with particular warmth the participation of the Bordeaux group on liquid crystals (Centre de Recherches Paul Pascal); of the theory group at the Ecole de Physique et Chimie (Paris); of T. Lubensky whose teaching and advice have been essential to both of us; of J. Toner, who convinced one of us (J. P.) to start the project (the central argument being: “you will be so happy when it is finished’); of a number of visitors and friends—V. Amarias, J. P. Carton, E. Dubois Violette, C. Garland, S. Lagerwall, L. Limat, S. Ramaswamy—who participated both in stimulating discussions and painful proofreadings. Last, but not least: our thanks to the late G. Berner; to M. Bertrand and G. Friedelance, who prepared the manuscript. And our nostalgic thanks to our companions, children and friends, who endured this long gestation. Paris LP. June 1993 P.G. de G.PREFACE TO THE FIRST EDITION Liquid crystals are beautiful and mysterious; I am fond of them for both reasons. My hope is that some readers of this book will feel the same attraction, help to solve the mysteries, and raise new questions. We have known of the existence of liquid crystals for a comparatively long time—eighty years—yet many experiments which could have been done thirty years ago are only now being performed. The importance of potential applications to thermography and electro-optic display was realized only ten years ago (mainly through the work of Fergason and Hellmeier), but the apparent lack of applications in earlier times cannot, by itself, explain certain delays. More fundamentally, the study of liquid crystals is complicated because it involves several different scientific disciplines: chemistry, optics, and mechanics, more specialized tools such as nuclear magnetic resonance, and also a certain sense of vision in three-dimensional space in order to visualize complex molecular arrangements. A semi-theorist like myself is not very well trained in any of these techniques. For this reason, this book is very incomplete; certain aspects (and particularly the chemical aspects) are reduced to a bare minimum. On the other hand, what a theorist can and should systematically introduce is comparisons with other fields. In the present context, comparisons with magnetic systems are often useful and will be mentioned. Comparisons between the so-called smectic phases and the superfluid helium II and superconductors are also highly instructive. How- ever, they imply a certain familiarity with low-temperature physics, which I did not want to impose as a prerequisite; for this reason, the references to superfluids are kept short. As do all theorists, I relish exact calculations, but in the present case I have tried to insert qualitative discussions rather than equations whenever Possible. Two parts remain particularly heavy and unpleasant, however: these are the sections of Chapters 3 and 5 concerned with the hydrostatics and hydrodynamics of nematics. This subject has been the source of some controversy between the classical mechanics group at Johns Hopkins and the theoretical physics group at Harvard. The two groups use very different languages; however (fortunately), their essential results are the same. I have tried to show this at the expense of a rather long and dull analysis, but, apart from this particular topic, the theoretical speculations have usually been condensed and replaced by one or two useful references. I must emphasize that the references quoted (both for experimental data and for theoretical work) do not claim to be a complete list and do not refer to historical priority. For instance, I do not mention the books of Lehmann and of Schenk because an average reader will get a much more clearviii PREFACE TO THE FIRST EDITION presentation of their main observations through the later reviews by Friedel or Saupet. Similarly, in the discussion of nematic order at the molecular level, I have omitted many theoretical papers which are interesting but still very tentative. Even as regards the choice of topics, the present work is incomplete; in particular, the lyotropic materials are not discussed. I believe that we still need a few years to understand them better. The same remark holds for the ‘exotic’ smectic phases D, E,F,G... and for the ‘blue phase’ of certain cholesterol esters. These phases are extremely interesting, but (in my opinion at least) we grasp them too poorly to discuss them in this book; my purpose is clarification more than compilation. Having taught the corresponding material during the past three years, I know that the present version is still very far from achieving this purpose, but I hope that, in spite of its obvious defects, it will help ‘liquid crystallers’ to reach a universally common language. My debt to a number of colleagues and friends is enormous. I would like to mention first G. Durand and M. Veyssié, who bravely started experiments on liquid crystals at Orsay during the autumn of 1968. Now, owing to their efforts and to the participation of other teams we have established a very active group; to a large extent, the present book represents its common views, based on constant discussions. The mistakes are my mistakes but the spirit is their spirit. I would also like to mention friends from outside the Orsay unit: J. Billard, who showed me the first liquid crystals that I ever saw and who, more recently, taught me a number of fundamental features of mesomorphic phases; R. B. Meyer, with whom I have been in constant correspondence and whom I hoped (in vain) to convince to write some chapters; our visitors from last year D. Litster, P. Pershan, and especially P. Martin, whose advice and positive criticism has been of constant help; F. C. Frank who, together with J. Friedel and M. Kleman, patiently instructed me on the structure of some difficult defects; G. Sarma and N. Boccara, with whom I have been exchanging ideas for a long time; S. Alexander, D. Martire, and J. Vieillard Baron, who coached me on the hard-rod problem. A special mention is due to J. L. Ericksen, H. Gruler, W. Marshall, A. Rapini, and Y. R. Shen for their close scrutiny of the initial manuscript, to C. Williams, who prepared the list of elastic constants, and to L. Léger, who did the same for the friction constants. Most of the typing work—particularly painful in view of the unpredictable changes of mood of the author—was done, with charming patience, by Marie France Jestin. M. Crasson and R. Seveste have succeeded in producing figures and photographs from my vague indications and scribblings. Last, but not least, I want to thank my wife Anne Marie for her + The history of liquid crystals has been analysed recently by H. Kelker (Molecular crystals and liquid crystals, 21, 1 (1973)).PREFACE TO THE FIRST EDITION ix cooperation: so many sunny weekends have been sacrificed to this book. Now, looking at the result, I am not entirely convinced that it was worthwhile, and I certainly do not dare to dedicate the book to her. ‘Well, now that we have seen each other,’ said the unicorn ‘If you believe in me, Til believe in you. Is that a bargain?” "Yes, if you like, said Alice. Orsay P.G. de G. December 19721 CONTENTS LIQUID CRYSTALS: MAIN TYPES AND PROPERTIES Lt 12 13 14 15 1.6 17 Introduction: what is a liquid crystal? The building blocks 1.2.1. Small elongated organic molecules 1.2.2 Small discoid organic molecules 1.2.3. Long helical rods 1.2.4 Polymers 1.2.5 Associated structures 1.2.6 Summary Nematics and cholesterics 1.3.1 Uniaxial nematics 1.3.2 Nematics of different symmetry 1.3.3 Cholesterics 1.3.4 Summary Smectics 1.4.1 Smectics A 14.2 Smectics C 1.4.3 ‘Hexatic’ smectics 1.4.4 ‘Crystalline’ smectics 1.4.5 The D phase Columnar phases 1.5.1 Hexagonal phases 1.5.2. Rectangular and oblique phases 1.5.3 ‘Antiphases’ More on long-, quasilong-, and short-range order 1.6.1 ‘Poor man’s’ elasticity of liquid crystals 1.6.2 Fluctuations 1.6.3 Long-range, quasilong-range, short-range order Remarkable features of liquid crystals LONG- AND SHORT-RANGE ORDER IN NEMATICS 21 2.2 Definition of an order parameter 2.1.1. Microscopic approach 2.1.2. Macroscopic approach 2.1.3 Relation between microscopic and macroscopic approaches Statistical theories of the nematic order 2.2.1 Mean-field calculations for hard rods and platelets 2.2.2. Mean-field theory with S? interaction 2.2.3. Computer calculations SOCMADWWH & 10 10 12 13 17 18 19 20 22 24 26 26 26 27 27 28 29 31 32 34 41 41 41 57 59 59 66 73CONTENTS 2.3. Phenomenological description of the nematic-isotropic transition 2.3.1 Landau free energy above T, 2.3.2 Static pretransitional effects 2.3.3 Biaxial phase 2.3.4 Failures of the mean-field approach 2.4 Mixtures 24.4 Importance of mixed systems 24.2 General trends STATIC DISTORTIONS IN A NEMATIC SINGLE CRYSTAL 3.1 Principles of the continuum theory 3.1.1 Long-range distortions 3.1.2 The distortion free energy 3.1.3 Discussion of the distortion-energy formula 3.1.4 Boundary effects 3.1.5 Nematics transmit torques 3.2 Magnetic field effects 3.2.1. Molecular diamagnetism 3.2.2 Definition of a magnetic coherence length 3.2.3. The Frederiks transition 3.3 Electric field effects in an insulating nematic 3.3.1 Dielectric anisotropy 3.3.2 Polarization induced by distortion (flexoelectric effect) 3.4 Fluctuations in the alignment 3.4.1 Light-scattering experiments 3.4.2 Orientation fluctuations and correlations in a nematic single crystal 3.4.3 Scattering of light by orientation fluctuations 3.5 Hydrostatics of nematics 3.5.1 Free energy and molecular field 3.5.2 Stresses and forces 3.5.3. The balance of torques DEFECTS AND TEXTURES IN NEMATICS 4.1 Observations 41.1 Black filaments (‘structures 4 fils’) 4.1.2 ‘Structures 4 noyaux’ (‘Schlieren structures’) 4.1.3 Types of defects 4.2 Disclination lines 4.2.1 Definition of the ‘strength’ 4.2.2 Distortion field around one line 4.2.3 The concept of line tension 4.3 Point disclinations 4.3.1 Instability theorem for lines of integral strength 716 76 8 82 85 91 91 92 98 98 98 100 102 108 116 117 117 120 123 133 133 135 139 139 141 144 150 150 151 155 163 163 163 163 166 168 168 170 178 180 18044 45 46 CONTENTS 4.3.2 Interpretation of the noyaux 4.3.3 Other observations on point defects Walls under magnetic fields 4.4.1 180° walls 4.4.2. Walls associated with a Frederiks transition 4.4.3. Transformation from walls to lines (‘pincement’) Umbilics Surface disclinations DYNAMICAL PROPERTIES OF NEMATICS 51 5.2 5.3 5.4 The equations of ‘nematodynamics’ 5.1.1 Coupling between orientation and flow 5.1.2 Choice of dynamical variables 5.1.3 The entropy source for a flowing nematic 5.1.4 The laws of friction Experiments measuring the Leslie coefficients 5.2.1. Laminar flow under a strong orienting field 5.2.2 Attenuation of ultrasonic shear waves 5.2.3 Laminar flow in the absence of external fields 5.2.4 Variable external fields 5.2.5 Inelastic scattering of light Convective instabilities under electric fields 5.3.1. Basic electrical parameters 5.3.2 Experimental observations at low frequencies 5.3.3. The Helfrich interpretation 5.3.4 Extension to higher frequencies Molecular motions 5.4.1 Dielectric relaxation 5.4.2 Nuclear spin-lattice relaxation 5.4.3. Acoustic relaxation 5.4.4 Translational motions 5.4.5 Temperature variation of the friction coefficients 5.4.6 Semi-slow motions above T, CHOLESTERICS 6.1 6.2 Optical properties of an ideal helix 6.1.1 The planar texture 6.1.2 Bragg reflections 6.1.3 Transmission properties at arbitrary wavelengths (normal incidence) 6.1.4 Interpretation 6.1.5 Conclusions and generalizations Agents influencing the pitch 6.2.1 Physicochemical factors 6.2.2 External fields xu 183 184 185 187 189 190 194 194 198 198 198 199 200 204 209 209 213 218 220 227 230 230 233 236 239 245 245 246 248 250 254 255 263 263 263 264 268 271 279 280 281 2866.3 6.4 6.5 CONTENTS Dynamical properties 6.3.1 Studies on small motions in a planar texture 6.3.2 Macroscopic flow 6.3.3 Convective instabilities 63.4 Torques induced by a heat flux Textures and defects in cholesterics 64.1 Textures 64.2 Singular lines Blue phases 6.5.1. Experimental observations 6.5.2 Double twist theories 6.5.3 Landau theory MACROSCOPIC BEHAVIOUR OF SMECTICS AND COLUMNAR PHASES TA 72 Continuum description of smectics and columnar phases: statics 7.1.1 Choice of variables 7.1.2 Distortion free energy of (non-chiral) smectics 7.1.3 Distortion free energy of columnar phases 7.1.4 Boundary conditions 71.5 Particular geometries 7.16 Transitions induced by external forces: Helfrich-Hurault effect 71.7 Transitions induced by external forces: undulation by mechanical tension 71.8 Transitions induced by external forces: thermo-optic effect in smectics A 71.9 Fluctuations Continuum description of chiral smectics and columnar phases 721 Chiral St and S* 72.2 Chiral Sf (S¥, Sf, and St) 7.2.3 Electric terms 724 Unwinding of the helix by an electric field 7.2.5 Fluctuations 7.2.6 Surface anchoring 7.2.1 Particular geometries DYNAMICAL PROPERTIES OF SMECTICS AND COLUMNAR PHASES 8.1 Unified description 8.1.1 Preliminary remarks 8.1.2 Basic equations 8.1.3. Smectics A and columnar hexagonal phases 294 295 296 308 313 320 320 323 329 337 337 337 341 349 353 357 361 364 369 370 376 376 377 379 385 389 393 408 408 408 410 415CONTENTS xv 8.1.4 The undulation mode 418 8.1.5 Permeation mode 420 8.1.6 Acoustic waves 421 8.1.7 Transverse modes 423 8.1.8 Sppex and Sc hydrodynamics 424 8.1.9 Remarks on the mode structure 429 8.2 Flow properties 430 8.2.1 Typical geometries 430 8.2.2 Flow past obstacles 435 8.2.3 Flow alignment 445 8.3 Breakdown of elasticity 452 8.3.1 Poor man’s derivation 452 8.3.2 Splay modulus 454 8.3.3 Macroscopic non-linear stress-strain relation 455 8.3.4 Renormalization group results 455 8.4 Breakdown of hydrodynamics 457 8.4.1 Poor man’s derivation 457 8.4.2. More general results 460 8.4.3 Columnar phases 461 8.4.4 Experimental situation 462 9 DEFECTS IN SMECTICS AND COLUMNAR PHASES 466 9.1 Observations 466 9.1.1 Large deformations in smectics 466 9.1.2 Large deformations in columnar phases 473 9.1.3 Dislocations 476 9.1.4 Disclinations 478 9.1.5 Walls 483 9.2 Dislocation and associated stress/strain 487 9.2.1 Distortions in smectics 487 9.2.2 Strain fields in columnar phases 500 10 PHASE TRANSITIONS IN SMECTICS 507 10.1 The A 2 N transition 507 10.1.1 Mean-field description 507 10.1.2 The analogy with superconductors 509 10.1.3 Critical phenomena 514 10.1.4 Anisotropic scaling S17 10.1.5 Experimental situation 518 10.1.6 The search for an anisotropic fixed point 521 10.1.7 Nature of the N-S, transition 522 10.1.8 ‘Poor man’s’ dislocation unbinding transition 524 10.1.9 Twisted grain boundary phase 525 10.1.10 Current situation 527xvi CONTENTS 10.2 Smectic A~smectic C transition 10.2.1 10.2.2 10.2.3 10.2.4 10.2.5 10.2.6 Superfluid analogy Ginzburg criterion Experimental findings First-order S, 2 Sc N-A-C point The S,-S¢ transition in two dimensions 10.3. Transitions involving hexatic phases 10.3.1 10.3.2 S, @ Sguex transition Harmonics of hexatic ordering and scaling properties 10.4 Frustrated smectics 10.4.1 10.4.2 10.4.3 10.4.4 10.4.5 10.4.6 Author index Subject index Experimental facts Frustrated smectics model Bilayer, partial bilayer smectics, incommensurate phases, and antiphases Re-entrant behaviour S,~S, isolated critical point and nematic bubble Molecular aspects 527 527 529 532 532 534 540 547 547 549 551 551 557 560 568 370 575 587 5951 LIQUID CRYSTALS: MAIN TYPES AND PROPERTIES “Que m’a donné le monde que ce mouvement d’herbes’ Saint John Perse 1.1 INTRODUCTION: WHAT IS A LIQUID CRYSTAL? During our years in high school, we have all been taught that matter only exists in three states: solid, liquid, and gas. However, this is not quite correct. In particular, certain organic materials do not show a single transition from solid to liquid, but rather a cascade of transitions involving new phases. The mechanical properties and the symmetry properties of these phases are intermediate between those of a liquid and those of a crystal. For this reason, they have often been called liquid crystals. A more proper name is ‘meso- morphic phases’ (mesomorphic: of intermediate form). To understand the significance of these new states of matter, it may be useful to recall first the distinction between a crystal and a liquid. In a crystal the components (molecules, or groups of molecules) are regularly stacked. The centres of gravity of the various groups are located on a three-dimensional periodic lattice. In the liquid the centres of gravity are not ordered in this sense. These two states of matter differ most obviously by their mechanical properties; a liquid flows easily. More fundamentally, a crystal is defined by the fact that, if a primitive pattern (or basis) is located at a point xg, the probability of finding an equivalent pattern at the point X = Xp + mA; + nA. + N33 (n; = integer; ie {1, 2, 3} and {a,} basis vectors) stays finite when |x — xo| > oo. As a result, its X-ray diffraction pattern shows sharp Bragg reflections characteristic of the lattice. In other words ; lim o(x)p(x')> = F(x — x’) ay where 0 The important feature is the existence of an (isotropic) length scale € over which correlations are lost. The X-ray diffraction pattern of a liquid exhibits diffuse peaks of typical width €~1. We are now in a position to define liquid crystals. These are systems in which a liquid-like order exists at least in one direction of space and in which some degree of anisotropy is present (a better definition of ‘some degree of anisotropy’ is: the density—density correlation function does not depend solely on the modulus |x — x’| but also on the orientation of x — x’ with respect to macroscopically defined axes). With this definition in mind we see that mesophases can be obtained in two different ways 1, Imposing no positional order, or imposing positional order in one or two rather than in three dimensions. All this can happen in nature. @ The first case obviously corresponds to a liquid but, if the correlation function is anisotropic, this is not an isotropic liquid; it is a nematic (there are at least two length scales ¢, and 2, over which correlations decay; || (respectively 1) indicate parallel (respectively perpendicular) to a macroscopically defined direction). @ The second case describes one-dimensional order in three dimensions: the system can be viewed as a set of two-dimensional liquid layers stacked on each other with a well defined spacing; the corresponding phases are called smectics. @ The third case corresponds to two-dimensionally ordered systems in three dimensions. They can be described as a two-dimensional array of liquid tubes and are called columnar phases. 2. Introducing degrees of freedom that are distinct from the localization of the centres of gravity. For non-spherical molecules, their orientation is the most natural candidate. Orientational transitions may take place in crystals, in liquid crystals, or in isotropic liquids. In the first two cases the freezing of orientational degrees of freedom does not generate very novel phases. On the other hand, freezing out orientational degrees of freedom in an isotropic liquid does yield anisotropy: certain organic liquids show a low-temperature phase in which the molecules are aligned preferentially along one direction. The correlations along the preferred direction are not equivalent to those in the perpendicular plane. The phase obeys our definition of a nematic: orientational freezing is the most transparent way of obtaining nematics.THE BUILDING BLOCKS 3 Nematics, smectics, and columnar phases are the only known forms of liquid crystals. We also have ‘colloidal crystals’, which can sometimes flow almost like water. But they are not liquid crystals: they are genuine crystals, exhibiting long-range three-dimensional order, with an unusually small yield stress. The type of liquid crystals that may be observed depends heavily on the structure of the constituent molecules or groups of molecules (the ‘building blocks’ as we shall call them). Nematics and smectics are often made of elongated objects, whereas some other nematics and most columnar phases are often made of disk-like molecules. Depending upon the nature of the building blocks and upon the external parameters (temperature, solvents, etc.), we can observe a wide variety of phenomena and transitions amongst liquid crystals. This chapter therefore starts off by discussing the building blocks that give rise to the known types of liquid crystals and then goes on to give a broad classification of nematics, smectics, and columnar phases. 12 THE BUILDING BLOCKS As explained above, to generate a liquid crystal one must use anisotropic objects, either elongated or disk-like. At the present time we know several ways to achieve this: with small molecules; with long helical rods that either occur in nature or can be made up artificially; with polymers; or with more complex units that are associated structures of molecules and ions. 1.2.1 Small elongated organic molecules The classical example is p-azoxyanisole (PAA) with the formula { oO From a (rough) steric point of view, this is a rigid rod of length ~20 A and width ~5 A. (The two benzene rings are nearly coplanar.) Another historical example is N-(p-methoxybenzylidene)-p-butylaniline (MMBA) with the formula cH, \ of \-ci=n CH, CH, \cy/ \ cHy ‘CH,4 LIQUID CRYSTALS: MAIN TYPES AND PROPERTIES Both PAA and MBBA are ‘nematogens’. This word means that they give tise to the nematic type of mesophase that will be discussed in Section 1.3.1. However, for PAA the nematic state is found only at high temperatures (between 116°C and 135°C at atmospheric pressure), while MBBA is nematic from ~20°C to 47°C, thus allowing much easier experimenta- tion. MBBA, however, lacks chemical stability. For many purposes cyano- biphenyls and cyanobicyclohexy] derivatives (and many others! ) are preferred nowadays. A broad class of organic molecules with the following general pattern are either nematogens or smectogens (ie. two parasubstituted aromatic rings linked by a large variety of functions preserving the long axis of the molecules, or one parasubstituted aromatic ting and one parasubstituted cyclohexane linked by the same type of functions; R and R’ can be polar or non-polar). The major types inside this scheme are listed in Table 1.1. Empirical rules describing the influence of the molecular nature on phase diagrams and related problems are reviewed in references 1-3. Smectics are favoured by long aliphatic chains, and amphiphilic interactions (ie. a tendency for the aliphatic and polarizable part to segregate). Another favourable class is obtained with cholesterol esters, of general formula 0 R—co~ (We use a simplified convention in which the hydrogens are not explicitly shown.) Note that the rings are not aromatic and the structure is not planar. However, the ring system is rigid, while the saturated chain C and the radical R (when it is not too short) behave like two somewhat more flexible tailsTHE BUILDING BLOCKS x{o)-x{0)- R’orR NEMATICS AND CHOLESTERICS 1.3.1. Uniaxial nematics A schematic representation of the order of a ‘nematic’} phase is shown in Fig. 1.4(a), (b). The main features are as follows. + The word ‘nematic’ was invented by G. Friedel. It comes from the Greek vnua = thread, and refers to certain thread-like defects that are commonly observed in these materials. The physical nature of these defects (‘disclination lines’) will be discussed in Chapter 4.NEMATICS AND CHOLESTERICS MW \ 11 CL Viy/ | ooé \ oO iY ao @ (b) Fig. 1.4. The arrangement of molecules in the nematic mesophase: (a) made up of rod-like molecules; (b) made up of disk-like molecules. 1. The centres of gravity of the molecules have no long-range order and, consequently, as already stated, there is only diffuse scattering in the X-ray diffraction pattern. The correlations in position between the centres of gravity of neighbouring molecules are similar to those existing in a conventional liquid, aside from the anisotropy 2, # ¢, in the length scale. In fact, nematics do flow like liquids. For a typical nematic such as PAA the viscosities are of order 0.1 poise.t 2. There is some order, however, in the direction of the molecules; they tend to be parallel to some common axis, labelled by a unit vector (or ‘director’) n. This is reflected in all macroscopic tensor properties: for instance, optically, a nematic is a uniaxial medium with the optical axis along n. (The difference between refractive indices measured with polariza- tions parallel or normal to n is quite large: typically 0.2 for PAA.) In all known cases, there appears to be complete rotational symmetry around the axis n. 3. The direction of n is arbitrary in space; in practice it is imposed by minor forces (such as the guiding effect of the walls of the container). This is a situation of broken rotational symmetry, reminiscent of a Heisenberg ferromagnet, where all spins tend to be parallel but where the energy is independent of the direction of the total moment M. 4. The states of director n and —n are indistinguishable. For instance, if the individual molecules carry a permanent electric dipole, as in Fig. 1.5(a), there are just as many dipoles ‘up’ as there are dipoles ‘down’ and the system is not ferroelectric. 5. Nematic phases occur only with materials that do not distinguish between Tight and left; either each constituent molecule must be identical to its mirror image (achiral) or, if it is not, the system must be a ‘racemic’ (1:1) mixture of the right- or left-handed species (we shall come back to this point in Section 1,33). From a crystallographic point of view, properties 2, 4, and 5 + For comparison, the viscosity of water at room temperature is ~ 107? poise.12 LIQUID CRYSTALS: MAIN TYPES AND PROPERTIES v 2 9 «8 Fig. 1.5. (a) In a nematic single crystal, if the molecules carry a dipole (represented by an arrow), there are as many dipoles ‘up’ as there are dipoles ‘down’. (b) If dipoles are forced to be ‘up’ the shape asymmetry of the molecules leads to a natural splay of the structure. may be summarized by the symbol D,,, in Schoenflies notation (co/mm in international notations). The dual aspects of a nematic phase (liquid-like but uniaxial) are exhibited most spectacularly in the nuclear magnetic resonance (NMR) spectrum, the uniaxial symmetry causes certain line splittings (which are absent in the conventional isotropic liquid phase). On the other hand, the lines are relatively narrow; this implies rapid molecular motions and is a natural consequence of the fluidity. Some applications of these NMR measurements will be discussed in Chapters 2 and 5. 1.3.2 Nematics of different symmetry 1. Could we have a uniaxial liquid made with polar molecules, all the dipoles being aligned (Fig. 1.5(b))? This would be a liquid ferroelectric. Unfortun- ately, a phase like this is unstable: because the head and the tail of each molecule are generally different in size, the whole structure must deform [15]. (We analyse these deformations in Chapter 3.) 2. Could we have a uniaxial liquid made with molecules that carry magnetic moments along their long axis, and generate a liquid ferromagnet? From a formal point of view, this is possible [15]: the ‘head’ and ‘tail’ of the molecule may now be chemically identical and no mechanical instability appears. The difficulty here is in the physics, not with the symmetry. The exchange couplings between neighbouring magnetic moments are weaker than the conventional interactions between molecules (steric, Van der Waals, etc.): thus liquid ferromagnets should be very rare. 3. Biaxial nematics can be expected with building blocks shaped in the form of long rectangles (Fig. 1.6). The long expected biaxial phase (Pa 2 n) m has first been observed in lyotropic systems: when certain delicate condi- tions are met, one can have micelles of the required shape, and they do build up a remarkable anisotropic liquid [16]. Note that it is, however,NEMATICS AND CHOLESTERICS 13 Fig. 1.6. Naive representation of a biaxial nematic: the long sides and short sides of the rectangles are on the average parallel to, respectively, the z and y directions. important to bear in mind that one can have an uniaxial phase with biaxial building blocks, and a biaxial phase with uniaxial building blocks! Biaxial nematics have also been observed in polymeric and low-molecular- weight thermotropic systems, as referred to in Chapter 2. 6 4. Exotic nematics. Nematics of higher symmetry, such as Da(S.m) (hexagonal), O,,(m 3 m) (cubic) and m 5 m (icosahedral symmetry) may seem somewhat exotic. The existence of a hexagonal nematic has been conjectured (between a usual D,,, nematic and a hexagonal columnar phase) [17]. Optically, a hexagonal phase is uniaxial just as that of a conventional nematic. Icosahedral nematics have been invoked in super- cooled liquids prior to the glass transition [18]. Another good candidate for generating an icosahedral nematic could be a melt of quasi-crystals: this new state of matter, obtained for instance by rapidly quenching samples of aluminium alloyed with manganese, iron, or chromium [19], is characterized by the existence of incommensurate lengths along icosa- hedral symmetry axes. The X-ray diffraction pattern reveals a dense distribution of Bragg spots that show the appropriate symmetry. Melting can well yield first a broadening of the peaks before losing the angular bond orientational order, which would be exactly that of the icosahedral nematic. 13.3 Cholesterics If we dissolve in a nematic liquid a molecule that is chiral (ie. different from its mirror image), we find that the structure undergoes a helical distortion.14 LIQUID CRYSTALS: MAIN TYPES AND PROPERTIES z Fig. 1.7. The arrangement of molecules in the cholesteric mesophase. (The successive planes have been drawn for convenience, but do not have any specific physical meaning). The same distortion is also found with pure cholesterol esters (which are also chiral). For this reason, the helical phase is called cholesteric. 1.3.3.1. The helical structure Locally, a cholesteric is very similar to a nematic material. Again the centres of gravity have no long-range order and the molecular orientation shows a preferred axis labelled by a director n. However, n is not constant in space. The preferred conformation, shown in Fig. 1.7, is helical. If we call the z-axis the helical axis, we have the following structure for n ny = COs(qoz + $) ny = Sin(goz + $) . (1.3) n, =0. Both the helical axis (z) and the value of ¢ are arbitrary; we see here another type of broken symmetry. The structure is periodic along z and (since the states m and —n are again equivalent) the spatial period L is equal to one-half of the pitch: _t \4olNEMATICS AND CHOLESTERICS 15 1 Cholesteryl nonanoate 74-76°C 2 Cholesteryl decanate 3 Cholesteryl nonanoate ‘Temperature (°C) 2 8 2 8 ‘Temperature (°C) 80} 75.0 B 74.6 76 74) 3500 4000 4500 5000 5500 6000 6500 7000 ‘Wavelength (A) Fig, 1.8, Relation between pitch and temperature in typical cholesterics. What is plotted horizontally is the optical wavelength for Bragg reflection on the helical structure, which is equal to the pitch multiplied by the refractive index A ~ 1.5 [20]. Typical values of L are in the 3000A range, ic. much larger than the molecular dimensions. Since L is comparable to an optical wavelength, the periodicity results in Bragg scattering of light beams. We shall discuss these optical effects in Chapter 6. “ Both the magnitude and sign of qo are meaningful. The sign distinguishes between right- and left-handed helices; a given sample at a given temperature always produces helices of the same sign. If we change the temperature 7, Qo changes (Fig. 1.8). In some particular cases qo(T) may even change sign at a particular temperature T*. This case is interesting. 1. For T = T*, the material is found to behave like a conventional nematic. 2. When we cross the temperature 7*, we find that the physical properties such as specific heats, etc. remain quite smooth. Properties 1 and 2 show that the local molecular arrangements are indeed very similar in the nematic and in the cholesteric state, as was first noted by G. Friedel [14]. In the published literature, the cholesteric helix is sometimes visualized as