Abstract
Diffusion can be used to infer the microscopic features of a system from the observation of its macroscopic dynamics. Brownian motion accurately describes many diffusive systems, but non-Brownian and nonergodic features are often observed on short timescales. Here, we trap a single ultracold caesium atom in a periodic potential and measure its diffusion. We engineer the particle-environment interaction to fully control motion over a broad range of diffusion constants and timescales. We use a powerful stroboscopic imaging method to detect single-particle trajectories and analyse both non-equilibrium diffusion properties and the approach to ergodicity. Whereas the variance and two-time correlation function exhibit apparent Brownian motion at all times, higher-order correlations reveal strong non-Brownian behaviour. We additionally observe the slow convergence of the exponential displacement distribution to a Gaussian and-unexpectedly-a much slower approach to ergodicity, in perfect agreement with an analytical continuous-time random-walk model. Our experimental system offers an ideal testbed for the detailed investigation of complex diffusion processes.
Original language | English |
---|---|
Pages (from-to) | 137-141 |
Number of pages | 5 |
Journal | Nature Physics |
Volume | 13 |
Issue number | 2 |
DOIs | |
Publication status | Published - 2017 Feb 1 |
Externally published | Yes |
ASJC Scopus subject areas
- Physics and Astronomy(all)