Abstract:
Artificial microscale heat engines are prototypical models to explore the mechanisms of energy transduction in a fluctuation-dominated regime(1,2). The heat engines realized so far on this scale have operated between thermal reservoirs, such that stochastic thermodynamics provides a precise framework for quantifying their performance(3-6). It remains to be seen whether these concepts readily carry over to situations where the reservoirs are out of equilibrium(7), a scenario of particular importance to the functioning of synthetic(8,9) and biological(10) microscale engines and motors. Here, we experimentally realize a micrometre-sized active Stirling engine by periodically cycling a colloidal particle in a time-varying optical potential across bacterial baths characterized by different degrees of activity. We find that the displacement statistics of the trapped particle becomes increasingly non-Gaussian with activity and contributes substantially to the overall power output and the effciency. Remarkably, even for engines with the same energy input, differences in non-Gaussianity of reservoir noise results in distinct performances. At high activities, the effciency of our engines surpasses the equilibrium saturation limit of Stirling effciency, the maximum effciency of a Stirling engine where the ratio of cold to hot reservoir temperatures is vanishingly small. Our experiments provide fundamental insights into the functioning of micromotors and engines operating out of equilibrium.