The Demon in a vacuum tube

Abstract

In the present paper, several issues concerning the second law of thermodynamics, Maxwell's demon and Landauer's principle are dealt with. I argue that if the demon and the system on which it operates without dissipation of external energy are made of atoms and molecules (gas, liquid or solid) in thermal equilibrium (whose behaviour is described by a canonical distribution), then the unavoidable reason why the demon cannot successfully operate resides in the ubiquity of thermal fluctuations and friction. Landauer's principle appears to be unnecessary. I also suggest that if the behaviour of the demon and the system on which it acts is not always describable by a canonical distribution, as would happen for instance with the ballistic motion of electrons at early stages of thermionic emission, then a successful working demon cannot be ruled out a priori. A critical review of two recent experiments on thermionic emission Maxwell's demons is also given.