Local time statistics and permeable barrier crossing: From Poisson to birth-death diffusion equations

Barrier crossing is a widespread phenomenon across natural and engineering systems. While abundant cross-disciplinary literature on the topic has emerged over the years, the stochastic underpinnings of the process have yet to be linked quantitatively to easily measurable observables. We bridge this gap by developing a microscopic representation of Brownian motion in the presence of permeable barriers that allows us to treat barriers with constant asymmetric permeabilities. Our approach relies upon reflected Brownian motion and on the crossing events being Poisson processes subordinated by the local time of the underlying motion at the barrier. Within this paradigm, we derive the exact expression for the distribution of the number of crossings and find an experimentally measurable statistical definition of permeability. We employ Feynman-Kac theory to derive and solve a set of governing birth-death diffusion equations and extend them to the case when barrier permeability is constant and asymmetric. As an application, we study a system of infinite, identical, and periodically placed asymmetric barriers for which we derive analytically effective transport parameters. This periodic arrangement induces an effective drift at long times whose magnitude depends on the difference in the permeability on either side of the barrier as well as on their absolute values. As the asymmetric permeabilities act akin to localized ratchet potentials that break spatial symmetry and detailed balance, the proposed arrangement of asymmetric barriers provides an example of a noise-induced drift without the need to time modulate any external force or create temporal correlations on the motion of a diffusing particle. By placing only one asymmetric barrier in a periodic domain, we also show the emergence of a nonequilibrium steady state.