Magnetoplasmons no sistema de elétrons superficiais sobre hélio líquido confinado em um canal parabólico

Abstract

A lot of experimental and theoretical work have been performed in order to understand the collective modes of two-dimensional electron system (2DES) confined to restricted geometries. In particular, the system of surface electrons on helium has provided a very convenient tool to study edge magnetoplasmon properties at the nondegenerate regime. Recently, magnetoplasma excitations in a single wire of electrons on suspended helium films were reported. In this thesis, we study the magnetoplasma vvaves in the nondegenerate 2DES laterally confined in a narrovv channel of width W by a self-consistent potential given, in the parabolic approximation, by Vy = m0Ω2 y2 / 2 , where Ω is the confinement frequency. First, we have showed that this is a very good approximation for the experimental conditions. We have considered, for the actual situation, a metallic gate away from the 2D layer at distance d. We calculate the magnetoplasmon modes for strong magnetic fields, coc / Cl » 1 and in the ultra quantum limit when only the lowest (spin-split) Landau levei is occupied, in particular, hcoc »kBT. For evaluation of the spectra and the spatial structure of the magnetoplasmons, we employ an approach, based on the random-phase approximation. The extension for this problem is not trivial because a new length scale (T = ^2kBT/ m0Q2 » t0 is introduced, where (0 is the magnetic length. We also include the effect of dissipation on the magnetoplasmon spectra using the integral equation for the charge density obtained in a Drude-like approach. We provide a detailed discussion of our theoretical findings with the experimental observations of the conductance peaks as a function of the magnetic fields for various values of the holding field and we relate the peaks to the magnetoplasmon modes. From a comparison between of our numerical calculations and the experimental results from Valkering and Heijden, we conclude that: i) for the fixed qx - (2tt/ P)n> where P is the channel perimeter, the frequency of any magnetoplasmon is 1/5 in agreement with experiment and theoretical calculation done by Sokolov and Studart; ii) the fundamental mode was not observed in the experiment, but the upper order modes; iii) despite the self-consistent calculation of the charge profile, our model is unable to explain the dependence of the mode spectra on the holding potential as found by Valkering and Heijden