How small can Maxwell's demon be? -- Lessons from autonomous electronic feedback models

TL;DR

This paper investigates the minimal size and efficiency of autonomous Maxwell's demon models using three-terminal electronic systems that emulate feedback control to transfer heat against a temperature gradient.

Contribution

It introduces autonomous three-terminal models that mimic external feedback loops, showing the minimal three-level system is most efficient for entropy manipulation.

Findings

Three-terminal models can emulate feedback control without external intervention.

Minimal three-level system optimally uses heat dissipation for entropy control.

Autonomous systems can achieve Maxwell demon-like behavior without bipartite structure.

Abstract

External piecewise-constant feedback control can modify energetic and entropic balances, allowing in extreme scenarios for Maxwell demon operational modes. Without specifying the actual implementation of external feedback loops, one can only partially quantify the additional contributions to entropy production. This is different in autonomously operating systems with internal feedback. Traditional (bipartite) autonomous systems can be divided into controller and a controlled subsystem, but also non-bipartite systems can accomplish the same task. We consider examples of autonomous three-terminal models that transfer heat mainly from a cold to a hot reservoir by dumping a small fraction of it to an ultra-cold (demon) reservoir, such that their coarse-grained dynamics resembles an external feedback loop. We find that the minimal three-level implementation is most efficient in utilizing heat dissipation to change the entropy balance of the effective controlled system.