Purity sieve for models with factorizable interactions

TL;DR

This paper investigates how purity evolves in systems with factorizable interactions, proposing a modified purity sieve criterion that better accounts for long-term purity behavior in pointer state selection.

Contribution

It introduces a new criterion for pointer state selection that considers longer-term purity evolution, improving upon the traditional short-time purity sieve approach.

Findings

Purity at very small times is maximized when the initial state is an eigenstate of the interaction operator.

The traditional purity sieve may not be reliable for selecting pointer states in all models.

The modified criterion successfully recovers known pointer states in quantum Brownian motion models.

Abstract

Evolution of purity in case of factorizable interaction between an open system and an environment is investigated. We derive a simple expression for purity decrease at the first instants of evolution (when purity is close to unity). We find that purity at very small times is maximal when an initial state of an open system coincides with one of the eigenstates of the interaction operator, no matter how weak the interaction is. On the other hand it is widely known that in general eigenstates of the interaction are not pointer states. Therefore the procedure of selecting pointer states by purity maximization (known as "purity sieve") should not rely on short-time purity behavior. We propose a modification of the purity sieve criterion which approximately takes into account purity evolution at longer times. As an example of its applicability we recover known results for pointer states of a particle undergoing quantum Brownian motion; we point out that the criterion is not applicable for some other models, however. It is argued that the proposed modified purity sieve may be used for selecting pointer states of a particle undergoing decoherence through collisions.