Collisional statistics of a stochastic single file

The anomalous diffusion of a single file of Brownian particles moving on a circle at a given temperature is characterized in terms of nearest-neighbor collisions. The time and the distance a particle diffuses (normally) between two successive collisions are computed numerically; their means, distributions, and correlation functions are determined for different values of the file parameters and reproduced analytically by means of simple phenomenological arguments. Most notably, the jump autocorrelation functions develop slow power-law tails. The ensuing impact representation of the single file dynamics suggests an alternate description of the single file diffusion as a geometrically constrained fluctuation mechanism.