Experimental Studies on the Size Dependence of Sliding Charge-Density Wave Phenomena

Available from UMI in association with The British Library. This thesis describes an experimental study of the electronic transport properties of the charge-density wave states in niobium triselenide and orthorhombic-tantalum trisulphide. The study concentrates on the size dependence of phenomena associated with the depinning and subsequent motion of the charge-density wave, and on phenomena specifically associated with small samples. The first part of the thesis describes the theoretical background to this research and reviews the main phenomena associated with sliding CDW conduction. After this a description is given of the technique I developed for mounting an array of closely spaced contacts onto very small samples. The results of the size dependence investigation are then presented and a model for CDW depinning and phase-slippage is introduced. Finally, the results of measurements on very small samples are described and also explained using the proposed model. The main results of the size dependence investigation are: (i) the threshold fields, for depinning the CDWs that form in NbSe_3, are inversely proportional to the transverse sample dimension; and (ii) the phase -slip voltage, necessary to convert the charge carried by the CDW back into normal carriers, is independent of sample size. The measurements on very small samples of o-TaS _3 revealed: (i) step-like changes in resistance with varying temperature and sub-threshold current; (ii) regions of negative differential resistance; and (iii) sharp a.c.-d.c. interference features, (comparable with those published for NbSe_3), which showed that the threshold field had an oscillatory dependence on the amplitude of the radio frequency field. From the size dependence of the threshold field, it is concluded that the sample surface provides a significant, if not dominant, source of pinning. A phase-dislocation model is introduced and used to explain the lack of any size dependence in the phase-slip voltage: in this model the phase-slip voltage represents the force required to make a phase-dislocation 'climb' rather than that required to nucleate it, as had been previously assumed. The mechanism of CDW depinning and motion is seen as a plastic deformation of the CDW 'lattice' due to the generation and 'gliding' of large dislocation loops just below the pinned surface. It is argued that the results obtained with small samples, and also CDW memory phenomena reported elsewhere, are also explicable within this model. Therefore, it is concluded that the phase-dislocation model is successful in describing, at least qualitatively, many of the observed CDW phenomena and it is suggested that the model offers a suitable framework on which to base further work.