quantum Liouville von neumann density matrix
# Quantum Liouville–von Neumann Density Matrix Kinetics and Open Quantum System Dynamics in GAAFET and CFET Nano-Device Architectures
---
## Executive Summary
The Liouville–von Neumann equation governs the evolution of the quantum density operator $\hat{
ho}(t)$ in both isolated and open quantum systems. In semiconductor nanodevices—particularly Gate-All-Around FETs (GAAfets) and Channel-All-Around FETs (CFETs) operating at sub-5 nm dimensions—quantum transport of carriers must be treated rigorously using density matrix formalism. This article provides complete derivations of the Liouville–von Neumann equation from first principles, develops the theory of open quantum systems via Lindblad master equations incorporating decoherence mechanisms (electron-phonon scattering, electron-impurity scattering, phase-breaking collisions), establishes the connection to non-equilibrium statistical mechanics, and demonstrates practical numerical implementation via density matrix time-evolution. Understanding quantum transport through density matrix kinetics is essential for predicting sub-threshold swing degradation, hot-carrier effects, and quantum interference phenomena that determine device performance at the ultimate scaling limit.
---
## Table of Contents
1. Introduction: Quantum Transport at the Nanoscale
2. The Density Operator: Fundamental Concepts and Interpretation
3. The Liouville–von Neumann Equation in Isolated Systems
4. Hilbert Space Structure and Operator Algebra
5. Open Quantum Systems: Introduction to Decoherence
6. Lindblad Master Equation: Derivation and Physical Interpretation
7. Decoherence Mechanisms: Electron-Phonon, Electron-Impurity, Phase-Breaking Collisions
8. Connection to Classical Transport: Semiclassical Limit and Wigner Functions
9. Quantum Transport in GAAFET/CFET Channels: 1D/2D Nanowires
10. Numerical Solution Methods: Runge-Kutta and Krylov Subspace
11. Python Implementation: Density Matrix Solver and Quantum Transport
12. Device Applications: Subthreshold Behavior and Quantum Interference
13. Experimental Validation and Emerging Measurements
14. References & Further Reading
---
## 1. Introduction: Quantum Transport at the Nanoscale
### 1.1 Breakdown of Classical Transport
Classical (Boltzmann) transport theory assumes:
- Particles follow well-defined trajectories in phase space ($\mathbf{r}$, $\mathbf{p}$).
- Collisions are local events with well-defined scattering rates.
- Wave-particle duality effects are negligible.
These assumptions fail when:
- Device dimensions approach the de Broglie wavelength: $\lambda_{dB} = h / \sqrt{2m E}$.
- For electrons at 1 eV in Si: $\lambda_{dB} \sim 1$ nm.
- For sub-5 nm devices: quantum effects dominate.
### 1.2 Quantum Mechanical Description
At nanoscale, electrons must be described as quantum waves with wavefunctions $\psi(\mathbf{r}, t)$ and density operators $\hat{
ho}(t)$ that encode all information about the quantum state.
Key quantum effects in nanodevices:
- Quantum confinement: Discrete energy levels (subbands) in narrow quantum wells.
- Quantum tunneling: Carriers tunnel through potential barriers instead of classically surmounting them.
- Quantum interference: Multiple paths through the device interfere constructively or destructively.
- Phase coherence: Carriers maintain quantum coherence over coherence length $L_\phi$.
### 1.3 Role of Decoherence
Real devices are not isolated systems. Interactions with the environment (phonons, impurities, defects) cause decoherence—loss of quantum coherence and phase information.
The interplay between quantum effects and decoherence determines transport and device characteristics.
---
## 2. The Density Operator: Fundamental Concepts and Interpretation
### 2.1 Pure vs. Mixed States
Pure state: A quantum system with known wavefunctio n $|\psi
angle$. The density operator is:
$$\hat{ ho} = |\psi angle\langle\psi|$$
Mixed state: A statistical ensemble of pure states $\{|\psi_k
angle, p_k\}$ where $p_k$ is the probability. The density operator is:
$$\hat{ ho} = \sum_k p_k |\psi_k angle\langle\psi_k|$$
Properties:
- Hermitian: $\hat{
ho}^\dagger = \hat{
ho}$
- Positive: $\langle\phi|\hat{
ho}|\phi
angle \geq 0$ for all $|\phi
angle$
- Trace normalized: $ ext{Tr}(\hat{
ho}) = 1$
- Idempotent (pure state): $\hat{
ho}^2 = \hat{
ho}$ iff $ ext{Tr}(\hat{
ho}^2) = 1$
- Non-idempotent (mixed state): $ ext{Tr}(\hat{
ho}^2) < 1$ iff mixing
### 2.2 Expectation Values and Observables
For an observable $\hat{A}$, the expectation value is:
$$\langle A angle = ext{Tr}(\hat{ ho} \hat{A})$$
This holds for both pure and mixed states:
- Pure: $\langle A
angle = \langle\psi|\hat{A}|\psi
angle$
- Mixed: $\langle A
angle = \sum_k p_k \langle\psi_k|\hat{A}|\psi_k
angle$
### 2.3 Basis Representations
In a chosen basis $\{|i
angle\}$ (e.g., energy eigenstates), the density matrix has elements:
$$ ho_{ij} = \langle i | \hat{ ho} | j angle$$
Diagonal elements $
ho_{ii}$: Probability of finding system in state $|i
angle$ (populations).
Off-diagonal elements $
ho_{ij}$ (i ≠ j): Quantum coherence between states $|i
angle$ and $|j
angle$ (coherences).
### 2.4 Purity and Entropy
The purity is $\mathcal{P} = ext{Tr}(\hat{
ho}^2)$:
- Pure state: $\mathcal{P} = 1$
- Maximally mixed (d-dimensional): $\mathcal{P} = 1/d$
The von Neumann entropy is:
$$S = - ext{Tr}(\hat{ ho} \ln \hat{ ho}) = -\sum_i \lambda_i \ln \lambda_i$$
where $\lambda_i$ are eigenvalues of $\hat{
ho}$.
---
## 3. The Liouville–von Neumann Equation in Isolated Systems
### 3.1 Derivation from Schrödinger Equation
The time-dependent Schrödinger equation for a state $|\psi(t)
angle$ is:
$$i\hbar \frac{\partial |\psi(t) angle}{\partial t} = \hat{H} |\psi(t) angle$$
For a density operator $\hat{
ho}(t) = \sum_k p_k |\psi_k(t)
angle\langle\psi_k(t)|$, differentiate:
$$\frac{\partial \hat{ ho}}{\partial t} = \sum_k p_k \left( \frac{\partial |\psi_k angle}{\partial t} \langle\psi_k| + |\psi_k angle \frac{\partial \langle\psi_k|}{\partial t} ight)$$
Substituting the Schrödinger equation and its adjoint:
$$\frac{\partial \hat{ ho}}{\partial t} = \sum_k p_k \left( -\frac{i}{\hbar}\hat{H}|\psi_k angle\langle\psi_k| + \frac{i}{\hbar}|\psi_k angle\langle\psi_k|\hat{H}^\dagger ight)$$
Since $\hat{H}$ is Hermitian ($\hat{H}^\dagger = \hat{H}$):
$$\boxed{\frac{\partial \hat{ ho}}{\partial t} = -\frac{i}{\hbar}[\hat{H}, \hat{ ho}]}$$
where $[\hat{H}, \hat{
ho}] = \hat{H}\hat{
ho} - \hat{
ho}\hat{H}$ is the commutator.
This is the Liouville–von Neumann equation, the quantum analog of the classical Liouville equation.
### 3.2 Hamiltonian for Quantum Transport: Single-Particle Picture
For a carrier (electron/hole) in a device, the Hamiltonian is:
$$\hat{H} = \hat{T} + \hat{V}$$
where:
- $\hat{T} = \frac{\hat{\mathbf{p}}^2}{2m^*}$ is the kinetic energy (momentum operator $\hat{\mathbf{p}} = -i\hbar
abla$)
- $\hat{V}(\mathbf{r})$ is the potential energy (band structure + applied gate voltage)
In a discrete tight-binding basis (e.g., lattice sites or molecular orbitals):
$$\hat{H} = \sum_i \epsilon_i |i angle\langle i| + \sum_{ij} t_{ij} |i angle\langle j|$$
where $\epsilon_i$ are on-site energies and $t_{ij}$ are tunneling matrix elements.
### 3.3 Energy Eigenbasis Representation
In the energy eigenbasis $\{\hat{H}|n
angle = E_n|n
angle\}$:
$$\frac{\partial ho_{mn}}{\partial t} = -\frac{i}{\hbar}(E_m - E_n) ho_{mn} = -i\omega_{mn} ho_{mn}$$
where $\omega_{mn} = (E_m - E_n)/\hbar$ is the transition frequency.
Solutions:
$$ ho_{mn}(t) = ho_{mn}(0) e^{-i\omega_{mn}t}$$
Diagonal elements ($m = n$) are constant: $\partial
ho_{nn}/\partial t = 0$ (populations conserved).
Off-diagonal elements ($m
eq n$) oscillate at the Bohr frequency $\omega_{mn}$.
---
## 4. Hilbert Space Structure and Operator Algebra
### 4.1 Hilbert Space for Many-Electron Systems
For $N$ electrons in a device, the Hilbert space has dimension $d = 2^{N imes ext{# orbitals}}$ (accounting for spin).
For practical devices:
- Single quantum dot: $d \sim 10$–100
- Few-electron nanowire: $d \sim 100$–10,000
- Larger devices: $d$ becomes prohibitively large; approximations needed
### 4.2 Fermi Antisymmetry and Slater Determinants
Electrons are fermions with antisymmetric wavefunctions. A many-body state is constructed as a Slater determinant:
$$|\Psi angle = \frac{1}{\sqrt{N!}} \det \begin{vmatrix} \psi_1(\mathbf{r}_1) & \psi_2(\mathbf{r}_1) & \cdots \\ \psi_1(\mathbf{r}_2) & \psi_2(\mathbf{r}_2) & \cdots \\ \vdots & \vdots & \ddots \end{vmatrix}$$
or in second-quantization notation:
$$|\Psi angle = \hat{c}_1^\dagger \hat{c}_2^\dagger \cdots \hat{c}_N^\dagger |0 angle$$
where $\hat{c}_i^\dagger$ creates an electron in orbital $i$.
### 4.3 Single-Particle Density Matrix (Reduced Density Matrix)
For a many-body system, the single-particle density matrix is obtained by tracing out all but one electron:
$$\gamma_{ij} = \langle \hat{c}_j^\dagger \hat{c}_i angle = ext{Tr}(\hat{ ho} \hat{c}_j^\dagger \hat{c}_i)$$
This encodes the correlation between creating an electron at orbital $i$ and annihilating one at orbital $j$.
For a non-interacting (mean-field) system: $\gamma_{ij}$ fully specifies the physical state.
---
## 5. Open Quantum Systems: Introduction to Decoherence
### 5.1 System vs. Environment Decomposition
A complete quantum system = Device (system of interest) + Surroundings (environment).
$$\hat{ ho}_{ ext{total}}(t) = \hat{ ho}_{ ext{sys}}(t) \otimes \hat{ ho}_{ ext{env}}(t) + ext{correlations}$$
Decoherence: Coupling to the environment causes quantum information to "leak" from the system into environmental degrees of freedom.
Observable consequences:
- Loss of off-diagonal density matrix elements ($
ho_{ij}$ for $i
eq j$ decay)
- Approach to classical behavior
- Entropy increase
### 5.2 Reduced Density Matrix
By tracing out the environment:
$$\hat{ ho}_{ ext{sys}}(t) = ext{Tr}_{ ext{env}}(\hat{ ho}_{ ext{total}}(t))$$
the dynamics of the system are no longer unitary; they are described by a master equation.
### 5.3 Decoherence Time Scales
Coherence time $ au_\phi$: Timescale over which quantum coherences decay.
For semiconductors:
- At room temperature: $ au_\phi \sim 1$–100 fs (very short due to strong phonon coupling)
- At cryogenic temperatures (< 10 K): $ au_\phi \sim 1$–100 ps (improved)
The coherence length is $L_\phi = v au_\phi$:
- Room temperature Si: $L_\phi \sim 1$–10 nm
- Cryogenic Si: $L_\phi \sim 100$ nm – 1 μm
For sub-5 nm devices, $L_\phi$ can be comparable to device dimensions, making coherence effects observable.
---
## 6. Lindblad Master Equation: Derivation and Physical Interpretation
### 6.1 Lindblad Form
The most general form of a Markovian (memory-less) master equation preserving quantum mechanics (complete positivity) is the Lindblad master equation:
$$\boxed{\frac{d\hat{ ho}}{dt} = -\frac{i}{\hbar}[\hat{H}, \hat{ ho}] + \sum_k \gamma_k \left( \hat{L}_k \hat{ ho} \hat{L}_k^\dagger - \frac{1}{2}\{\hat{L}_k^\dagger \hat{L}_k, \hat{ ho}\} ight)}$$
where:
- $\hat{H}$ is the system Hamiltonian
- $\hat{L}_k$ are Lindblad operators (jump operators)
- $\gamma_k$ are decoherence rates
- $\{\cdot, \cdot\}$ denotes the anticommutator
### 6.2 Physical Interpretation of Lindblad Terms
First term (unitary evolution): $-\frac{i}{\hbar}[\hat{H}, \hat{
ho}]$ represents coherent quantum dynamics (Liouville-von Neumann equation).
Second term (dissipation):
- $\hat{L}_k \hat{
ho} \hat{L}_k^\dagger$: "Jump" due to a particular dissipative process (e.g., phonon emission/absorption)
- $-\frac{1}{2}\{\hat{L}_k^\dagger \hat{L}_k, \hat{
ho}\}$: Normalization to conserve trace and positivity
### 6.3 Examples of Lindblad Operators
Electron-phonon scattering:
$$\hat{L}_{e-ph}^{nm} = \sqrt{\Gamma_{nm}} \hat{c}_n^\dagger \hat{c}_m$$
where $\Gamma_{nm}$ is the scattering rate from state $n$ to state $m$.
Decay (energy loss):
$$\hat{L}_{ ext{decay}} = \sqrt{\gamma} \hat{\sigma}_-$$
where $\hat{\sigma}_- = |0
angle\langle 1|$ is the lowering operator.
Dephasing (pure decoherence without decay):
$$\hat{L}_{ ext{dephase}} = \sqrt{\gamma_\phi} \hat{\sigma}_z$$
where $\hat{\sigma}_z = |0
angle\langle 0| - |1
angle\langle 1|$ is the population difference.
---
## 7. Decoherence Mechanisms: Electron-Phonon, Electron-Impurity, Phase-Breaking Collisions
### 7.1 Electron-Phonon Scattering (Inelastic)
Electron-phonon coupling causes energy exchange. The Hamiltonian is:
$$\hat{H}_{e-ph} = \sum_{q\lambda} g_{q\lambda} (\hat{c}_{n'}^\dagger \hat{c}_n \hat{a}_{q\lambda} + ext{h.c.})$$
where $\hat{a}_{q\lambda}$ are phonon creation/annihilation operators, $q$ is phonon wavevector, $\lambda$ is polarization.
The scattering rate (golden rule) is:
$$\Gamma_{n o n'} = \frac{2\pi}{\hbar} \sum_q |g_{q}|^2 (N_q + 1) \delta(E_n - E_{n'} - \hbar\omega_q) \quad ext{(emission)}$$
$$\Gamma_{n o n'} = \frac{2\pi}{\hbar} \sum_q |g_{q}|^2 N_q \delta(E_n - E_{n'} + \hbar\omega_q) \quad ext{(absorption)}$$
where $N_q = 1/(e^{\hbar\omega_q/k_B T} - 1)$ is the Bose-Einstein phonon population.
For electrons in Si at room temperature:
- Acoustic phonon scattering: $ au \sim 100$ fs
- Optical phonon scattering: $ au \sim 1$ ps (for high-energy electrons)
- Total: $1/ au_{ ext{total}} = 1/ au_{ ext{ac}} + 1/ au_{ ext{op}}$
### 7.2 Electron-Impurity Scattering (Elastic)
Ionized donors/acceptors scatter electrons elastically (no energy change, only momentum).
Scattering rate (Coulomb potential $V_{ ext{imp}} \propto 1/|\mathbf{r} - \mathbf{R}_{ ext{imp}}|$):
$$\Gamma_{n o n'} = \frac{2\pi}{\hbar} |V_{ ext{imp}}|^2 ho(E_F) \sin^2( heta/2)$$
where $ heta$ is scattering angle.
For Si at room temperature:
- Mobility-limited by impurity scattering: $\mu \sim 100$ cm²/Vs
- Scattering time: $ au \sim 200$ fs
### 7.3 Phase-Breaking Collisions (Inelastic)
Any inelastic process (energy-changing) breaks quantum coherence even if it doesn't change the carrier energy noticeably.
Example: A carrier in a superposition of states $|\psi
angle = (|A
angle + |B
angle)/\sqrt{2}$ interacts with a phonon. After interaction:
- State $|A
angle$ absorbed a phonon → orthogonal final state
- State $|B
angle$ emitted a phonon → different orthogonal final state
The interference pattern is destroyed; coherence is lost.
Phase-breaking time $ au_\phi \approx 2 au_{ ext{inelastic}}$ (Nyquist theorem).
### 7.4 Decoherence Time: Temperature Dependence
For Si, the coherence time scales as:
$$ au_\phi \sim \frac{\hbar}{k_B T} \quad ext{(rough scaling)}$$
- At 300 K: $ au_\phi \sim 1$ fs (estimated from $\hbar / k_B T \sim 10$ fs; actual measured $\sim 1$–10 fs)
- At 10 K: $ au_\phi \sim 100$ fs – 1 ps
- At 1 K: $ au_\phi \sim 1$ ps – 10 ps
The strong temperature dependence makes cryogenic operation essential for observing quantum effects in semiconductors.
---
## 8. Connection to Classical Transport: Semiclassical Limit and Wigner Functions
### 8.1 Wigner Function: Quasi-Probability Distribution
The Wigner function is a quasi-probability distribution in phase space that bridges quantum and classical mechanics:
$$W(\mathbf{r}, \mathbf{p}, t) = \frac{1}{(2\pi\hbar)^d} \int d^d \mathbf{s} \, e^{i\mathbf{p} \cdot \mathbf{s}/\hbar} \langle \mathbf{r} - \mathbf{s}/2 | \hat{ ho}(t) | \mathbf{r} + \mathbf{s}/2 angle$$
The Wigner function can be negative (unlike classical probabilities), but marginal distributions are always positive:
$$\int d^d \mathbf{p} W(\mathbf{r}, \mathbf{p}, t) = n(\mathbf{r}, t) \geq 0 \quad ext{(probability density in } \mathbf{r} ext{)}$$
$$\int d^d \mathbf{r} W(\mathbf{r}, \mathbf{p}, t) = u(\mathbf{p}, t) \geq 0 \quad ext{(probability density in } \mathbf{p} ext{)}$$
### 8.2 Wigner-Liouville Equation
The Wigner function evolves according to:
$$\frac{\partial W}{\partial t} + \mathbf{v} \cdot abla_\mathbf{r} W + \mathbf{F} \cdot abla_\mathbf{p} W = ext{quantum correction terms}$$
In the classical limit ($\hbar o 0$), the quantum correction terms vanish, and the Wigner-Liouville equation becomes the Boltzmann equation:
$$\frac{\partial f}{\partial t} + \mathbf{v} \cdot abla_\mathbf{r} f + \mathbf{F} \cdot abla_\mathbf{p} f = \left(\frac{\partial f}{\partial t} ight)_{ ext{coll}}$$
This shows that Boltzmann transport is the semiclassical limit of quantum density matrix transport.
### 8.3 Quantum Corrections to Conductivity
For weakly quantum systems (small $\hbar$ or large devices), quantum corrections to classical conductivity arise:
$$\sigma(\omega) = \sigma_0 + \Delta\sigma_{ ext{quantum}}(\omega)$$
Examples:
- Weak localization: Quantum interference enhances backscattering → conductance decreases
- Weak antilocalization: In materials with spin-orbit coupling, quantum interference reduces backscattering → conductance increases
For nanoscale devices at sub-5 nm: classical transport completely breaks down; full quantum treatment essential.
---
## 9. Quantum Transport in GAAFET/CFET Channels: 1D/2D Nanowires
### 9.1 GAAFET Architecture and Quantum Confinement
GAAFET (Gate-All-Around FET):
- Channel: Thin nanowire (~ 5–20 nm diameter) or nanosheet
- Gate: Surrounds all sides of channel
- Quantum confinement in cross-section (x-y) → discrete subbands
The channel is effectively a 1D quantum system with quantized transverse energy levels:
$$E_n = \frac{\hbar^2 \pi^2 n^2}{2m^* W^2}$$
where $W \sim 5$ nm is the nanowire width.
Energy spacing: $\Delta E \sim \frac{\hbar^2 \pi^2}{2m^* (5 ext{ nm})^2} \sim 50$ meV (comparable to thermal energy at room temperature).
### 9.2 Subband Structure and Transport
Transport occurs via discrete 1D subbands, each with its own density of states:
$$D_n(E) = \frac{m^*}{\pi \hbar^2 \sqrt{E - E_n}}$$ (1D van Hove singularity near band edge)
Quantum transport: Carriers travel as waves along the channel (longitudinal direction).
Conductance is quantized in units of $2e^2/h$:
$$G = \sum_n G_n = \sum_n \frac{2e^2}{h} T_n$$
where $T_n$ is the transmission through subband $n$ (0 or 1 for ballistic transport).
### 9.3 CFET (Channel-All-Around) Considerations
CFET (Channel-All-Around FET) is an extension where ALL four sides of a rectangular channel are gated (not just 3 sides like GAAFET):
- Further quantum confinement in both transverse directions (x and y)
- Multiple coupled 1D subbands → complex subband structure
- Enhanced gate control (all directions equally affected)
Density of states transitions from 1D to 0D (quantum dot-like) as confinement increases.
---
## 10. Numerical Solution Methods: Runge-Kutta and Krylov Subspace
### 10.1 Time-Evolution Operator
The Lindblad equation can be formally integrated:
$$\hat{ ho}(t) = \mathcal{U}(t, 0) \hat{ ho}(0) \mathcal{U}^\dagger(t, 0)$$
where $\mathcal{U}(t, t_0)$ is the time-evolution superoperator satisfying:
$$\frac{d}{dt}\mathcal{U}(t, t_0) = \mathcal{L} \mathcal{U}(t, t_0)$$
with superoperator $\mathcal{L}$ defined by the Lindblad equation.
### 10.2 Runge-Kutta Integration
For time-independent Hamiltonian and dissipation, use 4th-order Runge-Kutta:
$$\hat{ ho}(t + \Delta t) = \hat{ ho}(t) + \frac{\Delta t}{6}(k_1 + 2k_2 + 2k_3 + k_4)$$
where $k_i$ are intermediate steps computed via the Lindblad equation.
Advantages: Simple, stable, widely implemented.
Disadvantages: For large systems ($d > 10^4$), storage of $d imes d$ density matrix prohibitive; Hilbert space must be truncated.
### 10.3 Krylov Subspace Methods
For large systems, expand $\hat{
ho}(t)$ in a basis of orthogonal density matrices spanning a low-dimensional subspace:
$$\hat{ ho}(t) \approx \sum_k c_k(t) \hat{ ho}_k$$
The dimension can be reduced from $d^2$ (full density matrix) to $m \ll d^2$.
Time-dependent Variational Principle (TDVP): Variationally optimize the basis to minimize truncation error.
Matrix Product Density Operator (MPDO): For 1D systems, use tensor network representation analogous to Matrix Product State (MPS).
---
## 11. Python Implementation: Density Matrix Solver and Quantum Transport
"""
Quantum Liouville-von Neumann Density Matrix Solver
1D nanowire transport with electron-phonon scattering (Lindblad dissipation)
"""
import numpy as np
import matplotlib.pyplot as plt
from scipy.integrate import odeint
from scipy.linalg import expm
# Physical constants
hbar = 1.055e-34 # J·s
e_charge = 1.602e-19 # C
m_e = 9.109e-31 # kg
k_B = 1.381e-23 # J/K
# Quantum wire parameters
W = 5e-9 # Width (5 nm)
L_channel = 30e-9 # Channel length (30 nm)
T = 300 # Temperature (K)
m_eff = 0.25 * m_e # Effective mass (Si)
# Compute subband energies (particle-in-box)
def subband_energy(n, W):
"""Confinement energy for nth subband"""
return (hbar**2 * np.pi**2 * n**2) / (2 * m_eff * W**2)
# Use first 3 subbands
n_subbands = 3
E_subbands = np.array([subband_energy(n+1, W) for n in range(n_subbands)]) / e_charge # in eV
print(f"Quantum Wire Parameters:")
print(f" Width: {W*1e9} nm")
print(f" Subband energies (eV):")
for i, E in enumerate(E_subbands):
print(f" SB{i+1}: {E*1e3:.2f} meV")
# Hamiltonian in subband basis
# Diagonal: subband energies
# Off-diagonal: scattering (tunneling between subbands due to disorder)
def Hamiltonian_subbands():
"""
Tight-binding like Hamiltonian mixing subbands via disorder scattering
"""
H = np.diag(E_subbands * e_charge) # Diagonal subband energies
# Off-diagonal scattering elements (disorder-induced)
V_disorder = 1e-3 * e_charge # meV scale
for i in range(n_subbands):
for j in range(i+1, n_subbands):
H[i, j] = V_disorder
H[j, i] = V_disorder # Hermitian
return H
H = Hamiltonian_subbands()
# Electron-phonon dissipation (Lindblad operators)
# Simplified model: scattering rate proportional to temperature
def electron_phonon_rate(T):
"""Electron-phonon scattering rate"""
# Acoustic phonon scattering: Gamma ~ T
Gamma_0 = 1e12 # s^-1 (baseline at 300K)
return Gamma_0 * T / 300
Gamma_ep = electron_phonon_rate(T)
print(f"
Decoherence rates:")
print(f" E-phonon scattering rate: {Gamma_ep/1e12:.2f} × 10^12 s^-1")
# Lindblad operators: energy-loss terms
L_operators = []
L_rates = []
# Inter-subband scattering (energy-changing)
for i in range(n_subbands):
for j in range(n_subbands):
if i != j:
# Operator: |j⟩⟨i| (transition from i to j)
L = np.zeros((n_subbands, n_subbands), dtype=complex)
L[j, i] = 1.0
L_operators.append(L)
L_rates.append(Gamma_ep / (n_subbands - 1)) # Equal branching
# Dephasing (pure decoherence without energy loss)
Gamma_dephase = Gamma_ep * 0.1 # Smaller dephasing rate
for i in range(n_subbands):
for j in range(n_subbands):
if i != j:
# Dephasing operator: |i⟩⟨j|
L_dephase = np.zeros((n_subbands, n_subbands), dtype=complex)
L_dephase[i, j] = 1.0
L_operators.append(L_dephase)
L_rates.append(np.sqrt(Gamma_dephase))
# Initial density matrix (pure state: electron in ground state subband 0)
rho_0 = np.zeros((n_subbands, n_subbands), dtype=complex)
rho_0[0, 0] = 1.0 # |0⟩⟨0|
print(f"
Initial state: electron in SB1 (ground subband)")
# Lindblad master equation
def lindblad_eom(rho_vec, t):
"""
Time evolution according to Lindblad master equation
rho_vec: vectorized density matrix (column vector)
"""
# Reshape vector back to matrix
rho = rho_vec.reshape((n_subbands, n_subbands))
# Unitary evolution: -i/ℏ [H, ρ]
drho_unitary = -1j / hbar * (H @ rho - rho @ H)
# Dissipative evolution
drho_dissipation = np.zeros_like(rho)
for L, gamma in zip(L_operators, L_rates):
L_dag = L.conj().T
drho_dissipation += gamma * (L @ rho @ L_dag - 0.5 * (L_dag @ L @ rho + rho @ L_dag @ L))
# Total evolution
drho = drho_unitary + drho_dissipation
# Vectorize for ODE solver
return drho.flatten()
# Time integration
t_max = 100e-15 # 100 fs
t_points = np.linspace(0, t_max, 200)
# Solve ODE
rho_vec_initial = rho_0.flatten()
solution = odeint(lindblad_eom, rho_vec_initial, t_points)
# Extract populations and coherences
populations = np.zeros((len(t_points), n_subbands))
coherences = np.zeros((len(t_points), n_subbands-1), dtype=complex)
for idx, rho_vec in enumerate(solution):
rho = rho_vec.reshape((n_subbands, n_subbands))
# Populations (diagonal)
for i in range(n_subbands):
populations[idx, i] = np.real(rho[i, i])
# Coherences (off-diagonal)
for i in range(n_subbands - 1):
coherences[idx, i] = rho[i, i+1]
# Convert time to femtoseconds
t_fs = t_points * 1e15
# Plot results
fig, axes = plt.subplots(2, 2, figsize=(14, 10))
# Subplot 1: Populations vs time
ax1 = axes[0, 0]
for i in range(n_subbands):
ax1.plot(t_fs, populations[:, i], linewidth=2, label=f'SB{i+1}')
ax1.set_xlabel('Time (fs)')
ax1.set_ylabel('Population')
ax1.set_title('Subband Populations vs Time (Lindblad Dynamics)')
ax1.legend()
ax1.grid(alpha=0.3)
ax1.set_ylim([-0.05, 1.05])
# Subplot 2: Coherence decay
ax2 = axes[0, 1]
coherence_magnitude = np.abs(coherences)
for i in range(n_subbands - 1):
ax2.semilogy(t_fs, coherence_magnitude[:, i] + 1e-10, linewidth=2, label=f'$\
ho$_{0}{i+1}')
ax2.set_xlabel('Time (fs)')
ax2.set_ylabel('|Coherence| (log scale)')
ax2.set_title('Quantum Coherence Decay')
ax2.legend()
ax2.grid(alpha=0.3, which='both')
# Subplot 3: Purity Tr(ρ²)
ax3 = axes[1, 0]
purity = []
for rho_vec in solution:
rho = rho_vec.reshape((n_subbands, n_subbands))
purity.append(np.real(np.trace(rho @ rho)))
ax3.plot(t_fs, purity, 'r-', linewidth=2.5)
ax3.axhline(1.0, color='k', linestyle='--', alpha=0.3, label='Pure state')
ax3.axhline(1/n_subbands, color='k', linestyle='--', alpha=0.3, label='Maximally mixed')
ax3.set_xlabel('Time (fs)')
ax3.set_ylabel('Purity Tr(ρ²)')
ax3.set_title('Purity Decay (Decoherence)')
ax3.legend()
ax3.grid(alpha=0.3)
ax3.set_ylim([0, 1.1])
# Subplot 4: Von Neumann entropy
ax4 = axes[1, 1]
entropy = []
for rho_vec in solution:
rho = rho_vec.reshape((n_subbands, n_subbands))
# Compute eigenvalues
eigenvalues = np.linalg.eigvalsh(rho)
eigenvalues = eigenvalues[eigenvalues > 1e-10] # Remove numerical noise
# Von Neumann entropy: S = -Tr(ρ ln ρ) = -Σ λ_i ln(λ_i)
S_vn = -np.sum(eigenvalues * np.log(eigenvalues))
entropy.append(S_vn)
ax4.plot(t_fs, entropy, 'purple', linewidth=2.5)
ax4.set_xlabel('Time (fs)')
ax4.set_ylabel('Von Neumann Entropy')
ax4.set_title('Entropy Increase (Irreversibility)')
ax4.grid(alpha=0.3)
ax4.axhline(np.log(n_subbands), color='k', linestyle='--', alpha=0.3, label='Max entropy')
ax4.legend()
plt.tight_layout()
plt.savefig('quantum_liouville_von_neumann.png', dpi=150, bbox_inches='tight')
plt.show()
print(f"
{'='*70}")
print(f"Quantum Transport Simulation Results")
print(f"{'='*70}")
print(f"Final populations:")
for i in range(n_subbands):
print(f" SB{i+1}: {populations[-1, i]:.4f}")
print(f"
Initial purity: {purity[0]:.4f} (pure state)")
print(f"Final purity: {purity[-1]:.4f} (mixed state)")
print(f"
Final entropy: {entropy[-1]:.4f}")
print(f"Max entropy (for {n_subbands} states): {np.log(n_subbands):.4f}")
print(f"{'='*70}")---
## 12. Device Applications: Subthreshold Behavior and Quantum Interference
### 12.1 Subthreshold Swing and Quantum Effects
The subthreshold swing (SS) characterizes how steeply the transistor turns off:
$$SS = \frac{d V_g}{d(\log I_d)}$$
Theoretical minimum (classical): $SS_{\min} = \frac{k_B T}{e} \ln(10) \approx 60$ mV/dec at room temperature.
Quantum effects can improve or degrade SS:
- Ballistic transport (coherent): Can achieve SS < 60 mV/dec (quantum advantage)
- Quantum oscillations: Resonant tunneling can cause negative differential resistance (oscillating I-V)
- Tunnel FET advantage: Using band-to-band tunneling (quantum tunneling) enables SS < 60 mV/dec
### 12.2 Resonant Tunneling Diodes and Quantum Interference
In structures with multiple potential wells (double barriers), resonant tunneling occurs:
- Electrons tunnel through first barrier into quantum well
- If well energy matches incident electron energy, transmission $T = 1$ (resonance)
- Sharp transmission peak → sharp current peak → negative differential resistance
This is exploited in Resonant Tunneling Diodes (RTDs) for ultra-fast switches and oscillators.
### 12.3 Hot Carrier Effects and Ballistic Breakdown
In ultra-short channels (< 20 nm), carriers can become ballistic—traveling without scattering over the channel length.
Consequences:
- Carriers gain high kinetic energy from applied field (become "hot")
- Hot carriers can inject into dielectric (causing gate oxide degradation)
- Can cause impact ionization (secondary carrier generation)
Density matrix formalism predicts these effects through the energy distribution evolution.
---
## 13. Experimental Validation and Emerging Measurements
### 13.1 Coherence Measurements via Weak Localization
Weak localization experiment: Measure conductance oscillations as function of magnetic field. Oscillations indicate quantum coherence.
Coherence length extracted from oscillation period: $L_\phi = \sqrt{D au_\phi}$.
### 13.2 Scanning Tunneling Microscopy (STM) Spectroscopy
STM measures local density of states (LDOS) via tunneling current. Subband structure and resonances are visible in LDOS.
### 13.3 Time-Resolved Measurements
Femtosecond spectroscopy: Pump-probe experiments can directly measure dephasing times and coherence decay.
---
## 14. References & Further Reading
1. Breuer, H. P., & Petruccione, F. (2002). *The Theory of Open Quantum Systems*. Oxford University Press.
2. Lindblad, G. (1976). "On the Generators of Quantum Dynamical Semigroups." *Communications in Mathematical Physics*, 48, 119–130.
3. Zurek, W. H. (2003). "Decoherence and the Transition from Quantum to Classical." *Reviews of Modern Physics*, 75, 715–775.
4. Datta, S. (2005). *Quantum Transport: Atom to Transistor*. Cambridge University Press.
5. Ferry, D. K., Goodnick, S. M., & Bird, J. (2009). *Transport in Nanostructures* (2nd ed.). Cambridge University Press.
6. Caldirola, P., Lugiato, L., & Montaldi, E. (2007). "Dissipative Systems." *Scholarpedia*, 2(9), 4225.
7. Jauho, A. P., Wingreen, N. S., & Meir, Y. (1994). "Time-Dependent Transport through a Mesoscopic Conductor." *Physical Review B*, 50, 5528–5544.