Information flows in nanomachines

Abstract

Nanomachines are often interpreted as autonomous Maxwell's demons or information engines. This interpretation is appealing from a conceptual viewpoint but sometimes lacks a rigorous and precise formulation. Moreover, it is unclear whether this interpretation is truly useful for analyzing the performance of nanomachines or designing more efficient ones. The concepts of mutual information and information flows provide a quantitative and rigorous account of the role of information in physical systems. In this chapter, we present a pedagogical review of these concepts and apply them to a specific energy transducer at the nanoscale, which couples the tunnel transport of electrons through a carbon nanotube with its mechanical oscillations.