Using Ensemble Monte Carlo Methods to Evaluate Non-Equilibrium Green's Functions

TL;DR

This paper explores a particle-based approach to non-equilibrium Green's functions (NEGF) for quantum transport simulation in semiconductors, demonstrating potential computational efficiency improvements in high electric field conditions.

Contribution

It introduces a particle method for NEGF that leverages natural basis functions, showing preliminary results that suggest computational advantages over traditional methods.

Findings

Preliminary results for quantum transport in silicon at 300 K.

Particle approach offers potential computational efficiency gains.

Method suitable for high electric field conditions.

Abstract

The use of ensemble Monte Carlo (EMC) methods for the simulation of transport in semiconductor devices has become extensive over the past few decades. This method allows for simulation utilizing particles while addressing the full physics within the device, leaving the computational difficulties to the computer. More recently, the study of quantum mechanical effects within the devices, effects which also strongly affect the carrier transport itself, have become important. While particles have continued to be useful in quantum simulations using Wigner functions, interest in analytical solutions based upon the non-equilibrium Green's functions (NEGF) have become of greater interest in device simulation. While NEGF has been adopted by many commercial semiconductor, there remains considerable computational difficulty in this approach. Here, a particle approach to NEGF is discussed, and preliminary results presented illustrating the computational efficiency that remains with the use of particles. This approach adopts the natural basis functions for use in a high electric field and the preliminary results are obtained for quantum transport in Si at 300 K. This approach appears to offer significant advantages for the use of NEGF.