Questioning the validity of non-extensive thermodynamics for classical Hamiltonian systems

Abstract

We examine the non-extensive approach to the statistical mechanics of Hamiltonian systems with $H=T+V$ where $T$ is the classical kinetic energy. Our analysis starts from the basics of the formalism by applying the standard variational method for maximizing the entropy subject to the average energy and normalization constraints. The analytical results show (i) that the non-extensive thermodynamics formalism should be called into question to explain experimental results described by extended exponential distributions exhibiting long tails, i.e. $q$-exponentials with $q>1$, and (ii) that in the thermodynamic limit the theory is only consistent in the range $0\le q\le 1$ where the distribution has finite support, thus implying that configurations with e.g. energy above some limit have zero probability, which is at variance with the physics of systems in contact with a heat reservoir. We also discuss the ($q$-dependent) thermodynamic temperature and the generalized specific heat.

To appear in EuroPhysics Letters

References in corpus (2)

• Essential discreteness in generalized thermostatistics with non-logarithmic entropy
• Generalized Diffusion

Cited by in corpus (11)

• Influence of the interaction range on the thermostatistics of a classical many-body system
• Possible Divergences in Tsallis' Thermostatistics
• Long-range interactions, doubling measures and Tsallis entropy
• Ricci curvature, isoperimetry and a non-additive entropy
• Tsallis statistics generalization of non-equilibrium work relations
• Entropies from coarse-graining: convex polytopes vs. ellipsoids
• Canonical Ensemble in Non-extensive Statistical Mechanics
• Moduli of curve families and (quasi-)conformality of power-law entropies
• Toward a relative q-entropy
• Power-law entropies for continuous systems and generalized operations
• A classification of nonequilibrium steady states based on temperature correlations