Percolation transition in two dimensional electron gas: A quantum cellular automaton model

TL;DR

This paper introduces a quantum cellular automaton model to describe a disorder-driven percolation transition in two-dimensional electron gas, revealing a metal-insulator transition with critical behavior and experimental relevance.

Contribution

It develops a novel quantum cellular automaton framework combining quantum and semi-classical regimes to model electron localization and percolation in 2DEG.

Findings

Identifies a disorder-driven percolation transition in 2DEG.

Shows divergence of correlation length and power-law behavior at transition.

Reproduces experimental features like conductivity singularity and transition universality.

Abstract

A new type of disorder-driven electronic percolation transition is found for two-dimensional electron gas (2DEG), based on a quantum cellular automaton model. This transition is shown to be accompanied with a metal-insulator transition, as well as a singularity in the electronic compressibility. To this end, the electronic system which is assumed to be in contact with an electronic reservoir, is meshed by using of the phase coherence length $\zeta_{\phi}$ as an extent which divides the spatial dynamics of the electrons into two separate regimes and controls the localization of electrons. For the scales much smaller than $\zeta_{\phi}$ the treatment is quantum mechanical, whereas for the scales much larger than $\zeta_{\phi}$ the picture of semi-classical transport works and the classical Mote Carlo method is used. Thomas-Fermi-Dirac (TFD) theory is employed to find the dependence of the free energy of each cell on the temperature ($T$), the (charged) disorder strength ($\Delta$) and the charge content of the cell, and a cellular automaton model for transporting the electrons between the cells is developed. At the transition line (in the $T-\Delta$ space) the geometrical (e.g. correlation length) quantities also diverge and some power-law behaviors emerge with some critical exponents in agreement with the Gaussian free field (GFF) and the percolation theory. In the percolating side the spanning cluster probability (SCP) (as a realization of the conductivity) has a decreasing behavior in terms of the temperature which is the characteristics of the metallic phase. Our simple model yields the important features of the experimental observations, e.g. the singularity in the conductivity in some critical density and also the universality (non-universality) of the metal-insulator transition (MIT) for the small (large) disorders in 2DEG.