Reservoir model for twodimensional electron gases in quantizing magnetic fields: a review
Download 1.56 Mb. Pdf ko'rish
|
71 adabiyot zawadzki2013
Review Article
basic solid state physics b status solidi www.pss-b.com physica ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim LL degeneracy. The result of Baraff and Tsui was con firmed within the same model by a somewhat simpler calculation of Bok and Combescot [3]. However, at the same time a different line of thought prevailed, explaining the quantum Hall phenomenon by a localization of electron states within the LLs, see the reviews [4, 5]. The interpretation based on the electron localization became so dominant that during a certain period it was dif ficult to publish different points of view, cf. Refs. [6, 7]. Still, the reservoir model has kept appearing in the literature under different names in order to explain various observations on the 2DEGs: quantum transport [8 – 10], Fermi energy behaviour [11], cyclotron resonance (CR) [12, 13], interband photo-magneto-luminescence [14], magnetic susceptibilities [15], magneto-plasmon (MP) dispersion [16, 17], etc. In his well known book, Mahan [18] treats the localization and reservoir interpreta- tions of QHE on equal footing. Recently, the electron reservoir made a convincing reappearance in monolayer graphene [19]. Thus it seems that now, when the smoke of battles over the quantum transport in 2DEGs is not as thick as it used to be, it is a good time to write a review on the subject. The purpose of our paper is to collect and brie fly discuss publications suggesting the presence of electron reservoir in various experiments on 2DEGs. An important place is reserved for the quantum magneto-transport effects which started the whole discussion, but other phenomena are also presented. In fact, the latter are often more convincing because the charge transport is dif ficult to describe. It is hoped that our paper will stimulate additional investigations to clarify obscure points concerning this important problem. An effort has been made to quote all the relevant literature on the subject. 2 Constant electron density versus constant Fermi level In the following section we consider brie fly thermodynamic properties of 2DEGs in two limiting situations. The first is the standard case of a constant electron density in the QW: N ¼ const. In thermodynamic terms this situation represents a canonical ensemble. The second describes the case of a 2DEG in contact with an external reservoir that can ‘pin’ the Fermi level E F . This situation represents a grand canonical ensemble. In order to emphasize the main features and make calculations easier we consider an extreme case of a large reservoir that can completely fix the value of E F ¼ const. We contrast the two situations in order to make the following considerations understandable. Finally, we quote very brie fly results for a self-consistent calculation. This is done for historic reasons, since a similar calculation was performed by Baraff and Tsui, and also because it represents a realistic case realized in GaAs/GaAlAs hetrostructures. 2.1 Constant electron density We consider 2DEG of noninteracting electrons in a parabolic, spherical energy band at a finite temperature T in the presence of a quantizing magnetic field B parallel to the growth direction. The spin degeneracy is included but it is assumed that the spin- splitting factor g ¼ 0. QWs and superlattices based on GaAs satisfy quite well these assumptions if the exchange enhancement of the g value is neglected (see below). An incorporation of the spin splitting is straightforward. We assume further that only one electric subband is populated. The description is based on the work of Zawadzki and Lassnig [20]. The energetic DOS is taken in the form of a sum of Gaussian peaks r ðEÞ ¼ 1 2pL 2 X n;s ffiffiffi 2 p r 1 G exp 2 E l ns G 2 " # ; ð1Þ where L 2 ¼ ch/eB, l ns ¼ hv c (n þ 1/2) þ (1/2)m B g s, v c ¼ eB/m c is the cyclotron frequency, n and s ¼ 1 are the Landau and spin quantum numbers, respectively, and G is the broadening parameter assumed constant. Two features should be emphasized at this point. First, in addition to the Gaussian peak of DOS at each LL, there is a common factor B in front of total DOS. This means that, as B increases, each LL can contain more and more electrons. Second, according to the form assumed in Eq. (1), DOS between LLs is negligibly small if their separation hv c is distinctly larger than G. This situation is illustrated in Fig. 1. The electron density in cm 2 is N ¼ 1 2pL 2 X n;s ffiffiffi 2 p r 1 g Z 1 0 exp 2y 2 ns 1 þ exp z h ð Þ dz ; ð2Þ where y ns ¼ (z u ns )/g, z ¼ E/kT, h ¼ E F /kT, u ns ¼ l ns /kT, g ¼ G/kT are the reduced quantities. The filling factor of the system is de fined as n ¼ 2pL 2 N, denoting the number of Wlodek Zawadzki received his PhD in 1964 and since 1961 has been employed in the Polish Acad- emy of Sciences. He worked in the National Magnet Lab. MIT in Cambridge Mass. (1965 –1967) and at Ecole Normale Superieure in Paris (1974). He is the author of over 250 original papers in theoret- ical physics of semiconductors and relativistic quantum mechanics. He specialized in 3D and 2D systems of narrow-gap semiconductors (NGS) in the presence of a magnetic field and is author of the analogy between behavior of electrons in NGS and relativistic electrons in vacuum. Professor Zawadzki was awarded the Maria Sklodowska-Curie Prize, Polish State Prize, Smoluchowski Medal of the Polish Physical Society, and is Distinguished International Scholar of the University of Michigan. 248 W. Zawadzki et al.: Reservoir model for 2DEGs in quantizing magnetic fields ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.pss-b.com Download 1.56 Mb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling