Reservoir model for two-dimensional electron gases in quantizing magnetic fields: A review


  • We dedicate this work to the memory of our collaborator Stephane Bonifacie, who died suddenly of a heart attack at the age of 32. Stephane was an exceptionally competent and gentle person. He greatly contributed to the subject of this paper by doing decisive work on the magneto-photoluminescence in GaAs/GaAlAs asymmetric quantum wells reported in our Ref. [14]. Stephane's untimely departure has been an irreparable loss to everybody who knew him.


We collect and review works which treat two-dimensional electron gases (2DEGs) in quantum wells (QWs, mostly GaAs/GaAlAs heterostructures) in the presence of quantizing magnetic fields as open systems in contact with outside reservoirs. If a reservoir is sufficiently large, it pins the Fermi level to a certain energy. As a result, in a varying external magnetic field, the thermodynamic equilibrium will force oscillations of the electron density in and out of the QW. This leads to a number of physical phenomena in magneto-transport, interband and intraband magneto-optics, magnetization, magneto-plasma dispersion, etc. In particular, as first proposed by Baraff and Tsui, the density oscillations in and out of QW lead to plateaus in the integer quantum Hall effect (IQHE) at values observed in experiments. The gathered evidence, especially from magneto-optical investigations, allows us to conclude that, indeed, in most GaAs/GaAlAs hetrostructures one deals with open systems in which the electron density in QWs oscillates as the magnetic field varies. Relation of the density oscillations to other factors, such as electron localization, and their combined influence on the quantum transport in 2DEGs, is discussed. In particular, a validity of the classical formula for the Hall resistivity ρxy = B/Nec is considered. It is concluded that the density oscillations are not sufficient to be regarded as the only source of plateaus in IQHE, although such claims have been sometimes made in the past and present. Still, our general conclusion is that the reservoir approach should be included in various descriptions of 2DEGs in the presence of a magnetic field. An attempt has been made to quote all the relevant literature on the subject.