Invariance principle for the random Lorentz gas -- beyond the Boltzmann-Grad limit (Q2006401)

The Lorentz gas with randomly placed spherical hard core scatterers is considered in \(\mathbb{R}^d\). The invariance principle is proved under the kinetic (Boltzmann-Grad) limit and simultaneous diffusive scaling of particle trajectory. The simultaneity feature is the main novelty, if compared to seminal papers by \textit{G. Gallavotti} [``Divergencies and the approach to equilibrium in the Lorentz and the wind-tree models'', Phys. Rev. 185, No. 1, 308--322 (1969; \url{doi:10.1103/PhysRev.185.308})], \textit{H. Spohn} [Commun. Math. Phys. 60, 277--290 (1978; Zbl 0381.60099)], and \textit{C. Boldrighini} et al. [J. Stat. Phys. 32, 477--501 (1983; Zbl 0583.76092)]. The main ingredients of the argument are a coupling of a mechanical trajectory with the Markovian random flight process, and probabilistic and geometric controls on the efficiency of this coupling. An extended review of previous contributions to the topic (the invariance principle) is provided. Similar earlier results are mentioned [\textit{A. Ruzmaikina}, Commun. Math. Phys. 261, No. 2, 277--296 (2006; Zbl 1130.82313); \textit{E. Gasparim} and \textit{P. Ontaneda}, Commun. Math. Phys. 270, No. 1, 1--12 (2007; Zbl 1115.14006)] and substantial differences between the present work and the preceding research are outlined. In the present work the physical setting is different: low density instead of the weak coupling. The methodology is different as well, since probabilistic coupling replaces the analytic/perturbative one. Accordingly, the time scale of the validity of the diffusive approximation, while in terms of the kinetic time scale, is much longer and fully explicit. There are here two major approaches to the mathematically rigorous proof of the invariance principle. One refers to the quenched limit, another to the averaged-quenched (named also annealed) limit, with the averaging of the random initial velocity of the particle and the random placement of scatterers. The authors employ the annealing setting. Open problems are mentioned.