Memory loss is contagious in open quantum systems

TL;DR

This paper reveals that Markovian (memoryless) behavior in quantum systems can be transmitted between baths via the system, and introduces a new approach combining non-Hermitian Hamiltonians with master equations to better model these interactions.

Contribution

It uncovers the phenomenon of Markovianity being contagious between baths and develops a novel Bloch-Redfield-inspired method for more efficient quantum system-bath interaction modeling.

Findings

Markovianity can be transferred between baths through the system.

The new approach improves modeling accuracy for lossy systems interacting with non-Markovian baths.

Reduces computational complexity in quantum system-bath simulations.

Abstract

Memoryless (Markovian) system-bath interactions are of fundamental interest in physics. While typically, the absence of memory originates from the characteristics of the bath, here we demonstrate that it can result from the system becoming lossy due to the Markovian interaction with a second bath. This uncovers an interesting interplay between independent baths and suggests that Markovianity is ``contagious'', i.e., it can be transferred from one bath to another through the system with which they both interact. We introduce a Bloch-Redfield-inspired approach that accounts for this distinct origin of Markovianity and uniquely combines non-Hermitian Hamiltonian formalism with master equations. This method significantly improves the description of the interaction between a lossy system (associated with a Lindblad master equation) and a non-Markovian bath, reducing the computational demands of complex system-bath setups across various fields.