Stochastic entropy production for continuous measurements of an open quantum system

D. Matos, Lev Kantorovich, Ian J. FordJournal of Physics Communications 2022 · DOI

TL;DR

This paper investigates the "cost" of quantum measurement in terms of entropy production. By modeling a two-level quantum system interacting with its environment, the authors show that performing a measurement, even a weak continuous one, inherently generates entropy, reflecting the irreversible nature of the process. They demonstrate that this entropy production adheres to a detailed fluctuation theorem, providing a fundamental link between measurement and thermodynamics.

What Problem Does This Paper Address?

Quantum mechanics, at its core, describes systems that evolve deterministically. However, when a quantum system interacts with an environment, especially during a measurement, its behavior becomes inherently probabilistic and irreversible. This irreversibility is often quantified by entropy production. While classical thermodynamics has well-established ways to measure entropy production, applying these concepts to quantum systems, particularly during continuous measurements, presents significant theoretical challenges. Specifically, understanding the thermodynamic cost of a quantum measurement – how much "disorder" it generates – and how this relates to fundamental laws like the second law of thermodynamics, is a key question. This paper aims to bridge this gap by developing a framework to calculate stochastic entropy production for a quantum system undergoing continuous measurement.

Methods

The authors employ a combination of advanced theoretical tools:

  • Stochastic Liouville-von Neumann (SLN) Equation: This is the starting point, providing a non-Markovian description of an open quantum system. It's a stochastic differential equation that, when averaged, yields the deterministic evolution of the system's density matrix.
  • Markovian Approximation: To make the problem tractable and allow for a more direct interpretation of physical randomness, they take a Markovian limit of the SLN equation. This assumes the environment has no memory, simplifying the dynamics.
  • Quantum Trajectories (Unravelling): The core of their approach is to "unravel" the deterministic evolution into a set of stochastic trajectories. This is achieved by introducing stochastic terms that mimic the effect of a continuous measurement, allowing for the simulation of individual, physically realized measurement outcomes.
  • Stochastic Thermodynamics: They apply the principles of stochastic thermodynamics to these quantum trajectories. This involves calculating the entropy produced along each individual trajectory, rather than just considering the average behavior.
  • Fokker-Planck Equation: To analyze the probability distribution of the system's state variables over time, they derive and solve the associated Fokker-Planck equation. This is crucial for calculating the average entropy production.
  • Boundary Corrections: A subtle but important aspect is the inclusion of boundary corrections to the entropy production calculation, which are necessary due to singularities in the probability distribution at the boundaries of the system's state space.

Datasets

Not discussed in the paper. The study is theoretical and relies on simulations based on the derived equations.

Key Insights and Ideas

  • Measurement as an Irreversible Process: The paper strongly emphasizes that quantum measurement, even a weak continuous one, is fundamentally an irreversible process that generates entropy. This is not just a consequence of environmental interaction but a direct result of the act of measurement itself.
  • Stochasticity is Key: Averaging over all possible measurement outcomes (the deterministic approach) masks the true thermodynamic cost. Analyzing individual stochastic trajectories is essential to capture the entropy production associated with a specific measurement outcome.
  • Entropic Cost of Measurement: The authors quantify this cost, showing that connecting to and disconnecting from a measuring device both lead to positive mean entropy production. This suggests a fundamental thermodynamic price for gaining information about a quantum system.
  • Detailed Fluctuation Theorem Adherence: A significant finding is that the calculated stochastic entropy production satisfies a detailed fluctuation theorem. This theorem relates the probabilities of forward and reverse processes and is a powerful indicator of the underlying statistical mechanics.
  • Role of Coupling Strength: The strength of the coupling to the measuring device (represented by $\gamma$) significantly influences the system's dynamics and the resulting entropy production, pushing the system towards specific eigenstates.

Key Results

  • The developed framework successfully models continuous quantum measurements as stochastic processes.
  • They demonstrate that both connecting to and disconnecting a measuring device result in positive mean stochastic entropy production.
  • The entropy production for the process of disconnecting the measuring device is found to be higher than for the connection process.
  • The calculated entropy production distributions for different protocols show excellent agreement with the detailed fluctuation theorem, validating the theoretical framework.
  • The study highlights the necessity of boundary corrections for accurate entropy production calculations, especially when dealing with singular probability distributions.

Limitations and Future Work

  • High Temperature Approximation: The derivation relies on a high-temperature approximation ($\beta\epsilon \ll 1$). Relaxing this approximation would allow for the study of systems at lower temperatures.
  • Non-Markovian Dynamics: While the paper uses a Markovian limit for tractability, extending the framework to non-Markovian environments would provide a more general description.
  • More Complex Systems: The current study focuses on a two-level bosonic system. Future work could explore more complex quantum systems, including those with multiple energy levels or different types of environments.
  • Numerical Artefacts: The authors acknowledge subtle numerical artefacts in their calculations, which are addressed by analytical boundary corrections. Further refinement of numerical methods could mitigate these issues.
  • Generalization of Protocols: Exploring a wider range of measurement protocols and environmental couplings could reveal more about the fundamental relationship between quantum measurement and thermodynamics.

Summary updated

Abstract

Abstract We investigate the total stochastic entropy production of a two-level bosonic open quantum system under protocols of time dependent coupling to a harmonic environment. These processes are intended to represent the measurement of a system observable, and consequent selection of an eigenstate, whilst the system is also subjected to thermalising environmental noise. The entropy production depends on the evolution of the system variables and their probability density function, and is expressed through system and environmental contributions. The continuous stochastic dynamics of the open system is based on the Markovian approximation to the exact, noise-averaged stochastic Liouville-von Neumann equation, unravelled through the addition of stochastic environmental disturbance mimicking a measuring device. Under the thermalising influence of time independent coupling to the environment, the mean rate of entropy production vanishes asymptotically, indicating equilibrium. In contrast, a positive mean production of entropy as the system responds to time dependent coupling characterises the irreversibility of quantum measurement, and a comparison of its production for two coupling protocols, representing connection to and disconnection from the external measuring device, satisfies a detailed fluctuation theorem.
BibTeX
@article{matos2022stochastic,
  author = {D. Matos and Lev Kantorovich and Ian J. Ford},
  title = {{Stochastic entropy production for continuous measurements of an open quantum system}},
  journal = {Journal of Physics Communications},
  year = {2022},
  doi = {10.1088/2399-6528/aca742},
}
Figure