The Boltzmann equation is a foundational equation in statistical physics that describes how the velocity distribution of gas particles evolves over time due to collisions. A central concept here is entropy production, which measures how quickly a gas system is moving toward equilibrium. The H-theorem, proved by Boltzmann himself, guarantees that entropy production is always non-negative, meaning the system never spontaneously moves away from equilibrium overall. However, a natural follow-up question, posed by the mathematician McKean in 1966, is whether the rate of entropy production itself must decrease monotonically over time, or whether it could sometimes temporarily increase before eventually settling down.
This paper answers McKean's question with a clear "no." The authors construct explicit counterexamples showing that entropy production can actually increase for stretches of time before eventually decreasing. Crucially, these counterexamples work for two of the most physically important and well-studied cases: hard spheres, which model billiard-ball-like collisions relevant to real gases, and Maxwell molecules, which use a simplified mathematical model of particle interactions. The counterexamples are built from mixtures of Maxwellian distributions, meaning velocity distributions that individually look like equilibrium states but whose combination does not.
The significance of this result is that it closes a long-standing open question in kinetic theory and reveals that the approach to equilibrium in gas dynamics is more subtle than one might hope. While the system always trends toward equilibrium in the long run, the rate at which it does so can fluctuate non-monotonically. This means researchers cannot assume that entropy production serves as a simple, steadily decreasing quantity when analyzing how quickly or smoothly gases relax to their equilibrium states.