129Xe diffusion in a ferroelectric liquid crystal

Measurements of ¹²⁹Xe self-diffusion and shielding as a function of temperature were performed to cover the different phases of the ferroelectric liquid crystal FELIX-R&;D. The shielding data prove untwisting of the helical structure in the nematic phase (i.e. non-chiral nematic phase) of FELIX-R&;D. Self-diffusion measurements were carried out in a direction parallel to the main magnetic field of the NMR spectrometer. However, in order to yield the anisotropy of the xenon self-diffusion tensor a few measurements also were performed in the perpendicular direction. A special technique, based on the observation of the second spin echo instead of the conventional first one, was applied to avoid convection problems. The experiments reveal all the phase transitions and a continuous decrease in the self-diffusion constant along the external magnetic field, D‖, when moving from the isotropic to the smectic C? phase. The respective activation energy E‖ appears to vary remarkably, however, being about the same in the isotropic and smectic C? phases. In the smectic mesophases significantly faster xenon self-diffusion was detected in the perpendicular direction than in the parallel direction. The detected self-diffusion constant D ⊥ in the perpendicular direction seems to remain almost constant in the smectic mesophases and close to the value of the self-diffusion constant in the isotropic phase. The results are in agreement with the structural features of smectic phases and indicate redistribution of xenon atoms towards the interlayer space of smectic mesophases.     

Application of the Superconvergence Properties of the Galerkin Approximation to the Calculation of Upper and Lower Bounds for Linear Functionals of Solutions of Integral Equations

Superconvergence properties of the Galerkin approximation areassociated with recent methods developed for the calculationof upper and lower bounds of linear functionals of the solutionsof integral equations. For kernels with sufficiently smoothingproperties, the order of accuracy of prediction is four timesthe rate of convergence of the best approximation of the splinesubspaces in question. A simple example illustrates the theoreticalresults.