Feynman's ratchet established a principle: you cannot rectify thermal fluctuations into directed motion without either consuming energy or processing information. This is not a postulate — it follows from the second law. Brownian ratchets, molecular motors, Maxwell's demons — all require a driving force or an information-theoretic cost to produce net flow.
Jiang (arXiv:2603.01389) proposes a mechanism that appears to violate this constraint. Two liquid-vapor interfaces with different ion-induced structural asymmetries create an asymmetric potential energy landscape. Water molecules condensing at one interface release surface energy that is channeled by the structural asymmetry into persistent one-directional flow. No external energy input. No feedback mechanism. No information processing.
The mechanism relies on a distinction between equilibrium and the approach to equilibrium. The condensation process is spontaneous — surface energy is released as liquid forms from vapor. The structural asymmetry channels this released energy into net transport rather than dissipating it isotropically. The system is not in equilibrium (there is ongoing condensation), so the second law is not violated — the directed flow is powered by the free energy of condensation, which is a non-equilibrium process.
But this reframing raises the question: is “no external driving” the right description if the condensation itself is the driving force? The system is not at equilibrium. It has a free energy gradient (vapor → liquid). The structural asymmetry converts that gradient into directed flow. This is not noise rectification in the Feynman sense — it is free-energy-driven transport through an asymmetric channel. The subtlety is real, but the claim that thermal noise alone produces directed flow is stronger than the mechanism supports.
Whether the paper ultimately revises or merely clarifies the constraints on noise-driven transport depends on this distinction. The structure matters either way: asymmetry in the potential landscape converts isotropic energy release into anisotropic flow.
Jiang, “Self-sustained Molecular Rectification without External Driving or Information,” arXiv:2603.01389 (March 2026).