A path integral approach to macroscopic fluctuations out of equilibrium

by Mr Tommaso Cossetto (University of Luxembourg)

Thursday 15 Oct 2020, 13:00 → 14:00 Europe/Stockholm

Fysikum equipment

Statistical mechanics describes systems with a multitude of degrees of freedom at the price of stochasticity. Diffusive systems are usually described either with a Fokker-Plank equation, at the distribution level, or with a Langevin equation, at the trajectory level. A generalized Langevin equation can be written also for systems evolving with jump processes. I will use this as the starting point to build a path integral representation for the dynamics, related in a nontrivial way to the more known Doi-Peliti path integral. Then I will extend the construction to the generating function of time integrated observables. In the extensive limit, these tools give access to the macroscopic fluctuations of the observables and to their fluctuation relations. In the second part of the talk I focus on stochastic chemical networks. Non-linearity in the rate of chemical reactions can create a phase transition for the distribution of the steady current. Finally, I will describe how the macroscopic current responds to a perturbation of its driving force. Chemical networks of biological relevance can display negative differential response, which can be interpreted as a mechanism to optimize internal or external disturbances.

The seminar will be given on zoom: https://kth-se.zoom.us/j/69383379337