Well not literally The Maxwell's Demon... If a one-way reed valve can be brought down to size of only a few molecules, don't you think it could allow passage of a gas molecule across a barrier in one direction only? How I think it could work: -The small size (molecular) of the valve is critical to vastly reduce the probability of two gas molecules meeting at the valve at the same time from both sides of the barrier which will defeat the purpose the valve - a molecule coming in will also let molecule from other side out or prevent the valve from opening. -The one-way valve which may include a spring, must be lighter in weight than a single gas molecule. Either the valve must be made light or use a heavy gas or both. Example of gas that can be used is Tungsten or Sulfur Hexaflouride or the dangerous, Radon or Mercury Vapor. The reason for the valve having less weight is to prevent the gas in the intake side of the valve from being bounced back by the inertia of the valve. -The valves, probably millions/trillions of them could then surface an enclosing chamber. If this chamber is sealed, the pressure inside will keep increasing until the probability of both molecules from inside and outside arrives at the valve at the same time which either prevents the valve from opening or allows a molecule out after an arriving molecule comes in. -If the chamber can have a small a opening, a steady jet of gas is produced and work achieved, even in ambient (room) temperature. -Obviously, this chamber had to be enclose in a much larger and sealed gas chamber for safety purposes. Turbines and/or generators must be inside this sealed chamber. How it will not work: -Due to the molecular size of the valve, possibly only few atoms worth, will have elastic properties as well just like the gas molecules. A gas molecule forcing it open, will cause it to clap open and close indefinitely without loss of energy by heat/friction. The clapping action will allow gas molecules inside the chamber to escape. -I have no idea how to introduce friction/damping in the molecule valve so it won't clap. For best result, it must remain close once its returned to its resting position by the spring, until it's forced open by a gas molecule. Note that damping/friction is not really a significant loss in efficiency if the chamber can be thermally insulated. The action will heat the inside of the chamber which will return energy lost through damping back to the gas inside the chamber and ultimately, to the jet. -Has anyone made valves that small before or at least, lighter than the heaviest/densest known gas? What do you guys think?