A Rigorous Foundation for Stochastic Thermodynamics via the Microcanonical Ensemble

Xiangjun Xing 1,2,3 · 2025

Read A Rigorous Foundation for Stochastic Thermodynamics via the Microcanonical Ensemble by Xiangjun Xing 1,2,3 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-23604.txt.

Introduction

Xiangjun Xing builds strict foundations for stochastic thermodynamics in a 2025 paper. A small system, strongly gripped by a huge bath and driven from outside, exchanges heat in ways the field often defines loosely. Xing works with the joint system sealed off, counting states the classic way.

When the bath settles far faster than the driving, it stays in balance at each step. A split of the total energy makes the bath's share a slow invariant and defines the system's own energy cleanly. From time symmetry alone, the small system's motion proves memoryless with a local balance rule. Heat and bath entropy gain exact meanings, free of model detail.

Catalog record

CREATORXiangjun Xing 1,2,3
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE65 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-23604

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A Rigorous Foundation for Stochastic Thermodynamics via the Microcanonical Ensemble Xiangjun Xing 1,2,3 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 invoking any model details, we rigorously prove that the reduced dynamics of the system is Markovian and satisfies a form of local detailed balance (LDB) which involves transition probabilities but not path probabilities. By working entirely within the microcanonical framework and adopting a precise decomposition of the total energy, we provide rigorous definitions of bath entropy as the Boltzmann entropy, and of heat as the negative change of the bath energy. Our approach bypasses the ambiguities associated with conventional definitions of thermodynamic variables and path probabilities, and establishes a rigorous and thermodynamically consistent foundation for stochastic thermodynamics, valid even under strong system–bath coupling. I Introduction Stochastic thermodynamics Peliti-book ; sekimoto1998 ; Jarzynski-review ; Seifert-review provides a powerful framework for non-equilibrium physics of small systems coupled to fluctuating environments. Two cornerstones of this framework are the Markovianity and the local detailed balance (LDB). Markovianity allows us to ignore the detailed dynamics of the surrounding bath and to assign consistent probabilities to system trajectories. The LDB principle, in turn, provides the essential link between Markovian dynamics and the thermodynamics of the heat bath. Together, these two properties underpin entropy production seifert2005 , fluctuation theorems gallavotti1995 ; lebowitz1999 ; Crooks1999 ; Jarzynski1997 ; Bochkov1977 ; Bochkov1981 , and other major results in stochastic thermodynamics. The Markovian property of reduced dynamics is usually justified using projection operator techniques such as the Zwanzig–Mori formalism zwanzig1960 ; mori1965 ; Zwanzig-book , where weak coupling and near-equilibrium assumptions are introduced to facilitate explicit computations. However, in recent years, it has been gradually realized that these assumptions are not necessary to guarantee the emergence of Markovian dynamics Breuer-2005 ; Hanggi-2005 ; Xing-CG . The essential requirement is time-scale separation (TSS), which ensures that the bath remains in instantaneous equilibrium conditioned on the system’s state and control parameters. Despite this growing awareness, a model-independent derivation of Markovianity from TSS has remained lacking. The principle of LDB is usually presented as an extension of the equilibrium detailed balance (DB) condition for the transition probabilities bergmann1955 : P(x^{\prime}|x;t)\,p_{\text{eq}}(x)=P(x|x^{\prime};t)\,p_{\text{eq}}(x^{\prime}), (1) which, under weak coupling and bath equilibrium, can be recast into a form involving exchanged heat or change of bath entropy. Further assuming Markovian property (which ensures instantaneous bath equilibrium), this can be adapted to the LDB condition for non-stationary processes with time-dependent control parameter Crooks1999 : \frac{P_{F}[\gamma|\gamma_{0}]}{P_{B}[\tilde{\gamma}|\tilde{\gamma}_{0}]}=e^{-\beta Q[\gamma]}=e^{\Delta S_{Y}[\gamma]}, (2) where P_{F}[\gamma|\gamma_{0}] and P_{B}[\tilde{\gamma}|\tilde{\gamma}_{0}] are the conditional probabilities of a forward trajectory and its time-reversed counterpart, and Q[\gamma] is the heat transferred to the bath, whereas \Delta S_{Y}[\gamma] is the entropy change of the bath 1 1 1 Some author prefer an “unconditional” version of the LDB condition, in which both the forward and backward processes are assigned fixed initial distributions. While this approach is convenient for deriving certain fluctuation theorems, it introduces additional complications regarding the choice of initial conditions and the interpretation of trajectory ensembles, and somewhat obscures the essential new physics contained in LDB, which is the link between the system dynamics and the bath thermodynamics. We shall therefore focus on the conditional version of LDB in the present work. . While Eq. ( 2 ) has been widely adopted, it presumes weak interacton and Markovianity, that Q[\gamma] and S_{Y}[\gamma] are inferred from system dynamics, and that the path probabilities can be very subtle to define rigorously. Under strong coupling Jarzynski-2017-SC ; Hanggi-2020-comment ; Hanggi-2020-Colloquium ; Xing-SC-1 ; Xing-SC-2 or in the presence of multiplicative noises Xing-PI , Eq. ( 2 ) may lose precise meaning and can become dangerously misleading. A more formal derivation of LDB was supplied by Maes and Netocny Maes2003 ; Maes2020 , who relate the ratio of coarse-grained transition probabilities arising from an underlying time-reversible unitary dynamics to the difference in Boltzmann entropy (of the total system) between the final and initial states. A careful inspection indicates that these authors assumed the stationary nature of the dynamics. More importantly, their general formulation does not clarify the connection between Boltzmann entropy change and heat exchange between the system and its reservoirs, and thus remains detached from the concrete thermodynamic structure of physical systems. As Maes himself acknowledged in his later work Maes2020 , additional assumptions such as weak coupling and near-equilibrium conditions are required to apply the formalism to realistic systems. Hence the approach of Refs. Maes2003 ; Maes2020 shares the same limitations as the path-probability-based derivation of LDB due to Crooks Crooks1999 . At the heart of many ambiguities in stochastic thermodynamics lies the fact that the thermodynamics of the bath is not explicitly formulated—a direct consequence of the popular canonical ensemble approach. As a result, heat, bath entropy change, and entropy production must be inferred indirectly from the system dynamics, often leading to ambiguity or possible inconsistency. This ambiguity is substantially amplified if the coupling between the system and the bath is not fixed but needs to be tuned. For example, in Jarzynski’s original derivation of the nonequilibrium work relation, the system is assumed to be initially in thermal equilibrium with a heat bath, but the bath is then disconnected during the subsequent evolution in order to invoke Liouville’s theorem. This procedure is potentially dangerous: there is always legitimate doubt about subtle or uncontrolled effects that may arise when a bath is coupled or decoupled. While Jarzynski’s maneuver is mathematical in nature and may be circumvented by alternative methods of proof, there are physical processes—such as cyclic thermodynamic engines—where the tuning of the system-bath coupling is real and plays an essential role. In such cases, the ambiguities introduced by the canonical ensemble approach cannot be avoided. It is therefore conceptually advantageous to adopt the microcanonical ensemble, where the total system—including both the bath and the system—is modeled explicitly as an isolated Hamiltonian system, the bath thermodynamics is derived rather than assumed, and the system-bath coupling is treated explicitly. It is this more difficult and rarely pursued path that we shall follow in the present work, with the aim of establishing both Markovianity and LDB within a single, unified framework, irrespective of interaction strength and other model details. We note in passing that the microcanonical ensemble is also the natural ensemble for thermally isolated quantum systems, and has been extensively studied in the context of quantum fluctuation theorems and thermalization Talkner2008 ; Talkner2013 ; Polkovnikov2011 ; Goold2016 ; Vinjanampathy2016 ; Gogolin2016 . In contrast, our use of

Read the complete text