There are two paradigms to study nanoscale engines in stochastic and quantum thermodynamics. Autonomous models, which do not rely on any external time dependence, and models that make use of time-dependent control fields, often combined with dividing the control protocol into idealized strokes of a thermodynamic cycle. While the latter paradigm offers theoretical simplifications, its utility in practice has been questioned due to the involved approximations. Here, we bridge the two paradigms by constructing an autonomous model, which implements a thermodynamic cycle in a certain parameter regime. This effect is made possible by self-oscillations, realized in our model by the well-studied electron shuttling mechanism. Based on experimentally realistic values, we find that a thermodynamic cycle analysis for a single-electron working fluid is not justified, but a few-electron working fluid could suffice to justify it. Furthermore, additional open challenges remain to autonomously implement the more studied Carnot and Otto cycles.

• Received 13 January 2021
• Accepted 25 March 2021

DOI:https://doi.org/10.1103/PhysRevLett.126.180605