Electron-lattice coupling in conjugated polymers

The possibility of a number of applications has fuelled the study of conjugated polymers for some years. Because of the level of electron-lattice coupling in these materials, it is necessary to perform self-consistent calculations of electronic wave functions and atomic positions in order properly to study the defects responsible for their behaviour.Early empirical methods for performing such self-consistent calculations have proved remarkably successful in many of their predictions, but have still displayed certain shortcomings, primarily due to their failure satisfactorily to treat electron-electron interaction. They are still useful in a number of cases, however, and have been used here to derive some results which have not been obtained elsewhere, or which have only been derived in a more time-consuming manner than is necessary.This thesis is primarily concerned, though, with the development of a method which enables semi-empirical Hartree-Fock theory to be used in a self-consistent manner. The method is novel in that it performs the atomic relaxation and wave function calculation simultaneously and requires no numerical derivatives to be calculated. Both of these features lead to a method that is extremely efficient in terms of computer time and which can therefore reasonably be applied to molecules sufficiently large to model conjugated polymers.The results obtained by this new method are shown to be able to account for most of the shortcomings of the earlier methods, in particular their failure satisfactorily to explain the quenching of luminescence in cis-polyacetylene and their poor predictions of the relative strengths of the two photoinduced absorption peaks in polythiophene. The ability of trans-polyacetylene (t-PA) to support a novel type of dynamic defect known as a breather is also verified. A quantitative estimate is made of the mobility of the fundamental defect in t-PA, known as a soliton, and this is in good agreement with experiment.