arXiv:2610.18553
On the Increase of Boltzmann Entropy in Closed Macroscopic Quantum Systems
Cedric Igelspacher, Roderich Tumulka
Abstract
We consider a closed macroscopic quantum system in a pure state $ψ$ evolving unitarily with Hamiltonian $H$ in a finite-dimensional Hilbert space $\mathscr{H}$. Following von Neumann, we assume a decomposition of $\mathscr{H}$ into mutually orthogonal subspaces $\mathscr{H}_ν$, where each $ν$ represents a macro state, and $\mathscr{H}_ν$ contains the pure states compatible with $ν$. We define the quantum Boltzmann entropy for a macro state $ν$ as $S_\mathrm{qB}(ν) = k_{\mathrm{B}} \log \dim \mathscr{H}_ν$, where $k_\mathrm{B}$ is the Boltzmann constant. Goldstein et al. [http://arxiv.org/abs/1903.11870] have suggested, as a version of the second law of thermodynamics, that $S_\mathrm{qB}$ predominantly increases with time for realistic macroscopic systems. Here, we discuss precise versions of this statement and confirm them for several simple models of $H$ and $\mathscr{H}_ν$.
AI Summary
The paper asks how to formulate entropy increase for a closed quantum system in a pure state, using quantum Boltzmann entropy $S_{\mathrm{qB}}(\nu)=k_{\mathrm B}\log d_\nu$ for orthogonal macrospaces of dimensions $d_\nu$. Since unitary evolution generally produces superpositions of macrostates, it proposes two criteria: a weak second law requiring macrostate probabilities to move approximately toward higher entropy in stochastic order, and a strong second law requiring probability currents to flow predominantly from smaller to larger macrospaces. The strong criterion implies the weak one; violations are quantified by cumulative backward versus forward probability transport. Three toy-model settings are examined. For a small macrospace with uniformly random orientation, its occupation is approximated by the squared Fourier transform of the density of states. A theorem gives sufficient positivity and convexity conditions on that density for monotonic decay, including truncated power laws with exponent at least three. For GUE and Wigner Hamiltonians, existing random-matrix estimates yield explicit Bessel-function formulas for occupations and currents, with arbitrarily small pointwise errors at times $0<t\ll D^{1/7}$. These formulas suggest cumulative backward transport below approximately $2.5\%$ when an equilibrium macrospace dominates. Previously established results for a three-macrospace block-band model imply approximately monotonic relaxation without ultrafast thermalization. The random-orientation and Wigner models have physically unrealistic ultrafast relaxation. The $2.5\%$ conclusion is not established rigorously: simultaneous-in-time and accumulated-error control remain unresolved. Conditions guaranteeing either criterion for realistic macroscopic Hamiltonians are left open.
Error found: The Wigner-model cumulative-current conclusion is not established by the supplied fixed-time estimates without additional integration and macro-state-count control, as the appendix itself acknowledges. The stated 2.5% cumulative-violation bound neglects transitions into macro spaces smaller than the initial one and needs a stronger quantitative equilibrium-dominance assumption. Example 1 incorrectly extends its continuum approximation to times below an assumed exp(D) recurrence scale, although an allowed equally spaced spectrum recurs exactly at time 2πD.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 14 pages, 3 figures; v2: Theorem 1 added