Search

The dynamics of a globally neutral system of diffusing Coulomb charges in two dimensions, driven by an applied electric field, is studied in a wide temperature range around the Berezinskii-Kosterlitz-Thouless transition. I argue that the commonly accepted ``free particle drift'' mechanism of charge transport in this system is limited to relatively low particle densities. For higher densities, I propose a modified picture involving collective ``partner transfer'' between bound pairs. The new picture provides a natural explanation for recent experimental and numerical findings which deviate from standard theory. It also clarifies the origin of dynamical scaling in this context., 4 pages, RevTeX, 2 eps figures included; some typos corrected, final version to be published in Phys. Rev. Lett

We study the two-dimensional Ginzburg-Landau model of a neutral superfluid in the vicinity of the vortex unbinding transition. The model is mapped onto an effective interacting vortex gas by a systematic perturbative elimination of all fluctuating degrees of freedom (amplitude {\em and} phase of the order parameter field) except the vortex positions. In the Coulomb gas descriptions derived previously in the literature, thermal amplitude fluctuations were neglected altogether. We argue that, if one includes the latter, the vortices still form a two- dimensional Coulomb gas, but the vortex fugacity can be substantially raised. Under the assumption that Minnhagen's generic phase diagram of the two- dimensional Coulomb gas is correct, our results then point to a first order transition rather than a Kosterlitz-Thouless transition, provided the Ginzburg-Landau correlation length is large enough in units of a microscopic cutoff length for fluctuations. The experimental relevance of these results is briefly discussed. [Submitted to J. Stat. Phys.], 36 pages, LaTeX, 6 figures upon request, UATP2-DB1-93

We investigate superconducting pair correlations in the attractive Hubbard model on a finite square lattice. Our aim is to understand the pronounced size dependence which they display in the weak and intermediate coupling regimes. These size effects originate from the electronic shell structure of finite systems and severely complicate a reliable extrapolation of numerical simulation data from small systems to the thermodynamic limit. To analyze the size effects in detail, we use the BCS approximation, as well as a particle number conserving modification of it and compare the results with those of quantum Monte Carlo simulations. As an application, we explore the possibility of reducing the shell effects in simulation data by changing the shape of the system and the imposed boundary conditions and by making use of the size dependence of corresponding BCS data.

Using the BCS approach and quantum Monte Carlo simulations, we investigate the size dependence of the pairing correlations in the weak and intermediate coupling regimes. Pronounced size effects appear as long as the average spacing of the single-particle energy levels is larger than or comparable to the characteristic energy scale (gap). The irregularity of the size dependence is traced to variations in the occupation of the highest nonempty energy level with system size. We also propose means to reduce the size dependence and its irregularity.

We study the dynamics of a classical, two-component plasma in two dimensions, in the vicinity of the Kosterlitz-Thouless (KT) transition where the system passes from a dielectric low-temperature phase (consisting of bound pairs) to a conducting phase. We use two ``complementary'' analytical approaches and compare to simulations. The conventional, ``intuitive'' approach is built on the KT picture of independently relaxing, bound pairs. A more formal approach, working with Mori projected dynamic correlation functions, avoids to assume the pair picture from the start. We discuss successes and failures of both approaches, and suggest a way to combine the advantages of both., 4 pages, LaTeX, 4 eps figures included; to appear in the proceedings of the ``1999 International Conference on Strongly Coupled Coulomb Systems'', 4-11 Sept 1999 in Saint-Malo (France); some typos corrected