Liquid crystals: a chemical physicist's view

This paper has allowed me the rare opportunity and pleasure of acknowledging those who have played important roles in my scientific career. It has also enabled me to report work in the field of liquid crystals which has gone unpublished. The particular topics have been selected because they illustrate the areas of liquid crystal science with which I have been especially concerned. The predicted phase diagram of mixtures of rods and spheres is both intricate and interesting. The ability to test these predictions experimentally has required the use of quasi-spherical solutes such as tetraethyltin. The reasons for the failure of the experiments to conform to theory are investigated and explained in terms of the orientational order of this flexible molecule, determined using deuterium NMR spectroscopy. The tetrapodes are more exotic tetrahedral structures in which four mesogenic groups are linked to a central atom or group. The massive flexibility of such molecules poses a severe problem for the prediction of their liquid crystal behaviour. Here a solution to this problem is presented and used to predict the dependence of the transitional properties of the tetrapodes on the spacer length and the mode of its attachment to the mesogenic group. The same molecular field approach has been employed to predict the variation of the transitional properties of liquid crystal dimers with the length of the spacer. It is found that for spacers containing about 12 or more atoms the odd-even effect predicted for the transitional properties varies significantly depending on the model used to describe the spacer conformation. That is, whether the torsional angles defining the conformations are taken to be discrete or continuous. Cyanobiphenyl dimers with spacers containing up to 24 atoms have been synthesized to test these predictions. The Gay-Berne potential has proved to be an important model with which to study liquid crystal behaviour using simulation techniques. By joining two Gay-Berne particles together with a flexible ethane link we have constructed a Gay-Berne dimer and have been able to explore the properties of this mesogen. In particular its phase behaviour, the novel structure of the smectic A phase and how the conformational distribution alters with the phase have been studied. Despite its attractive features there are relatively simple systems for which the Gay-Berne potential is not suitable. These include molecules with a spherocylindrical shape and those with a sphere embedded at the centre of such a structure. In fact the shapes of many mesogenic molecules approximate to the former, and certain metallomesogenic molecules have shapes like the latter. The novel Corner S-function potential provides a valuable way to represent such cylindrically symmetric shapes and we use this to simulate the behaviour of these systems. It is found that the sphere has a major influence on the phase behaviour as well as on the crystal structure.