Harvesting information to control non-equilibrium states of active matter

TL;DR

This paper demonstrates how correlated noise can be used to control and induce transitions between non-equilibrium states in an optically trapped particle, mimicking active matter and revealing a link to information theory.

Contribution

It introduces a novel method to control active-like matter using correlated noise, enabling precise manipulation of non-equilibrium states and exploring the thermodynamics of information harvesting.

Findings

Transitions between non-equilibrium states can be triggered without work, only calorific cost.

Heat production is proportional to the spectral entropy of the noise.

The method mimics active matter and links to Landauer's principle.

Abstract

We propose to use a correlated noise bath to drive an optically trapped Brownian particle that mimics active biological matter. Thanks to the flexibility and precision of our setup, we are able to control the different parameters that drive the stochastic motion of the particle with unprecedented accuracy, thus reaching strongly correlated regimes that are not easily accessible with real active matter. In particular, by using the correlation time (i.e., the "color") of the noise as a control parameter, we can trigger transitions between two non-equilibrium steady states with no expended work, but only a calorific cost. Remarkably, the measured heat production is directly proportional to the spectral entropy of the correlated noise, in a fashion that is reminiscent of Landauer's principle. Our procedure can be viewed as a method for harvesting information from the active fluctuations.