Quenching Bloch oscillations in a strongly correlated material

TL;DR

This paper extends dynamical mean-field theory to analyze nonequilibrium Bloch oscillations in a strongly correlated material, revealing how these oscillations decay and develop beats under strong fields and interactions.

Contribution

It introduces a highly efficient, scalable computational method to study nonequilibrium dynamics in strongly correlated systems using the Falicov-Kimball model.

Findings

Beats with period 2π/U develop at strong fields

Bloch oscillations are sharply damped with strong interactions

Oscillations become irregular over time

Abstract

Dynamical mean-field theory is generalized to solve the nonequilibrium Keldysh boundary problem: a system is started in equilibrium at a temperature T=0.1, a uniform electric field is turned on at t=0, and the system is monitored as it approaches the steady state. The focus here is on the Bloch oscillations of the current and how they decay after their initial appearance near t=0. The system is evolved out to the longest time allowed by our computational resources--in most cases we are unable to reach the steady state. The strongly correlated material is modeled by the spinless Falicov-Kimball model at half-filling on a hypercubic lattice in d=oo dimensions, which has a Mott-like metal-insulator transition at U=sqrt{2}. The computational algorithm employed is highly efficient, parallelizes well, and scales to thousands of processors. For strong fields, we find beats develop with a period of 2 pi/U, while for strong interactions, the Bloch oscillations are sharply damped and become quite irregular in time.