Thesis Title: Lemuel John F. Sese
Thesis Title: Lemuel John F. Sese
Thesis Title: Lemuel John F. Sese
Department of Physics
Mapúa University
Thesis Title
Abstract
A two-dimensional elecron gas (2DEG) under a perpendicular magnetic field is considered. For
a given Landau level, the chemical potential can be approximated to have a linear dependence
on the magnetic field. This approximation is valid except when complete filling is reached
(integer filling factor). The chemical potential oscillation yields a recursion relation for every
Landau level. This gives way to an exact form of the magnetization. The derivations herein
hold at low temperature.
1
Contents
1 Objectives and significance 3
1.1 Two-dimensional electron gas in perpendicular magnetic field . . . . . . . . . . . 3
1.2 Chemical potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Magnetization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3 Timetable 5
2
1 Objectives and significance
Electron systems have their importance in technology mostly in semiconductors. These sys-
tems are well explained by quantum mechanics since electrons are quantum particles which are
describable by a wavefunction Ψ(~r, t). Despite of this, there are some phenomena that are un-
explainable when introducing some constraints. The example of these contraints are reducing
the dimension of a system, and is subjected to low temperature T and high magnetic field B.
Reducing the dimension from 3D to 2D to 1D will change the behavior of the system because
of their different density of states (DOS) [Kittel]. The DOS D(E) for 3D has a form of E 1/2
dependence, for 2D with a constant, and for 1D with a E −1/2 dependence. From these DOS,
we can solve the total internal energy of the system,
Z
U = dE ED(E)f (E) , (1)
where f (E) = 1/(1 + exp[(E − µ)/kB T ]) is the Fermi-Dirac distribution function, µ is the
chemical potential, and kB is the Boltzmann’s constant. Also, other thermodynamic properties
like µ can be obtained its behavior from the DOS by knowing the electron concentration,
Z
N = dE D(E)f (E) . (2)
There are numerous studies [Zawadzki, Villagonzalo, Li, Gammag, Zhu, Wang] in determin-
ing the behavior of different thermodynamic quantities by utilizing (1) and (2). One example
[Gammag] is by determining the behavior of µ using (2) by numerical simulation.
where e is the electron charge, h is Planck’s constant and En = (n + 1/2)~ωc is the nth Landau
level with units of energy. Here, ωc = eB/m∗ is the cyclotron frequency for a given effective mass
m∗ . If a system has a very narrow distribution of energy [Villagonzalo], then a delta-shaped
DOS can be used to model actual materials [Wilde].
For real systems, the actual shape of the density of states is determined by making theoretical
fits to the heat capacity data from experimental measurements [Gornik]. many experimental
and theoretical studies on the thermodynamic and magnetic properties of 2DEG, therefore, use
a from of DOS based on a Gaussian function [Ando, Smith, Zawadzki, Zawadzki(Surf. Sci.)]
such as
eB X 1 1/2 1 (E − En )2
D(E, B) = exp − , (4)
h n 2π Γ 2Γ 2
3
Figure 1: The magnetic field dependence of the chemical potential.
where Γ is the broadening parameter in units of energy. Here, Γ describes different types of
interaction such as electron-electron and electron-impurity scattering as describe in Ref. [Ando,
Yang]. The broadening parameter determines the width of distribution at the peaks of energies
En .
1.3 Magnetization
Considering the 2DEG as a grand canonical ensemble, its free energy [Greiner]
Z +∞
µ(B) − E
F = µ(B)N − kB T D(E, B) ln 1 + exp dE , (6)
0 kB T
where kB is the Boltzmann constant. If we substitute the ideal DOS (3) to (6), leads to
eB X
µ(B) − En (B)
F = µ(B)N − · kB T ln 1 + exp . (7)
h n
kB T
4
As can be seen from (7), the quantization of energy simplifies the calculation for the free energy.
From Maxwell’s relations, M can be obtained from the free energy F,
∂F
M (B, T ) = − . (8)
∂B N =constant
Applying (8) to (7),the derivative will act on the continuous region of µ and the magnetization
takes the form
∂µ eX
M (B, T ) = −N + (kB T β(B, n) + Bσ(B, n)) , (9)
∂B h n
where
µ(B) − En (B)
β(B, n) = ln 1 + exp (10)
kB T
and
∂µ 1 e~
σ(B, n) = fn (B) − n+ . (11)
∂B 2 m∗
Here, fn in (11) is the Fermi-Dirac distribution function,
1
fn (B) = h i . (12)
En (B)−µ(B,n)
exp kB T +1
3 Timetable