Comparing real-time coupled cluster methods through simulation of collective Rabi oscillations

TL;DR

This paper compares two time-dependent coupled cluster methods by simulating Rabi oscillations in multi-atom systems, revealing their stability, accuracy, and limitations in modeling collective quantum dynamics.

Contribution

It introduces a unified theoretical framework for TD-EOM-CC and TDCC, explaining their different scaling and stability behaviors in simulating collective Rabi oscillations.

Findings

TD-EOM-CC is numerically stable but unreasonably scales with system size.

TDCC captures initial energy scaling but breaks down near complete population inversion.

A shifted Hamiltonian framework explains the different behaviors of the two methods.

Abstract

The time-dependent equation-of-motion coupled cluster (TD-EOM-CC) and time-dependent coupled cluster (TDCC) methods are compared by simulating Rabi oscillations for different numbers of non-interacting atoms in a classical electromagnetic field. While the TD-EOM-CC simulations are numerically stable, the oscillating time-dependent energy scales unreasonably with the number of subsystems resonant with the field. The TDCC simulations give the correct scaling of the time-dependent energy in the initial stages of the Rabi cycle, but the numerical solution breaks down when the multi-atom system approaches complete population inversion. We present a general theoretical framework in which the two methods can be described, where the cluster amplitude time derivatives are taken as auxiliary conditions, leading to a shifted time-dependent Hamiltonian matrix. In this framework, TDCC has a shifted Hamiltonian with a block upper triangular structure, explaining the correct scaling properties of the method.