A Rigorous Foundation for Stochastic Thermodynamics via the Microcanonical Ensemble
Xiangjun Xing (Wilczek Quantum Center, School of Physics, Astronomy, Shanghai Jiao Tong University, Shanghai 200240, China)

TL;DR
This paper establishes a rigorous, microcanonical-based foundation for stochastic thermodynamics in Hamiltonian systems strongly coupled to a bath, proving Markovian dynamics and local detailed balance without model-specific assumptions.
Contribution
It introduces a microcanonical framework with a specific energy decomposition, rigorously deriving Markovian reduced dynamics and local detailed balance in strongly coupled Hamiltonian systems.
Findings
Bath energy is an adiabatic invariant under slow evolution.
Reduced system dynamics are Markovian and satisfy local detailed balance.
Provides rigorous definitions of heat and entropy within the microcanonical ensemble.
Abstract
We consider a small Hamiltonian system strongly interacting with a much larger Hamiltonian system (the bath), while being driven by both a time-dependent control parameter and non-conservative forces. The joint system is assumed to be thermally isolated. Under the assumption of time-scale separation (TSS)--where the bath equilibrates much faster than the system and the external driving--the bath remains in instantaneous equilibrium, described by the microcanonical ensemble conditioned on the system state and the control parameter. We identify a decomposition of the total Hamiltonian that renders the bath energy an adiabatic invariant under slow evolution. This same decomposition defines the system Hamiltonian as the Hamiltonian of mean force, and ensures that neither the system nor the control parameter does reactive work on the bath. Using time-reversal symmetry and TSS, and without…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsStatistical Mechanics and Entropy · Complex Systems and Time Series Analysis · Opinion Dynamics and Social Influence
