I propose you the following conjecture: is it possible to convert directly solar energy in a mechanical form?

Imagine a thin sheet metal made in a strong material like steel, with a thickness approaching 50 µm. This sheet metal is drilled by a very great number of conical holes, the small opening of holes on a face of sheet, the large opening on the other face. The opening angle of these conical holes is very small, approximately 0.1 degree. The diameter of the small opening of holes must be very small, lower or equal to the mean free path of the gas in contact. For example, the mean free path of the molecules of atmosphere at sea level is 70 nanometers. If the gas in contact is the atmosphere at sea level, the diameter of the small opening must be lower than 70 nm, the diameter of the other opening can be bigger, approaching some hundreds nanometers.

The kinetic theory of gases makes the assumption that the collisions of gas particles with the solid walls in contact are perfectly elastic. The smallness of the diameter of conical hole in the part close to the small opening makes that the molecules of gas contained in the hole have more frequently shocks on the wall of hole that among them. The slope of the wall of hole modifies gradually the trajectories of gas molecules. Each elastic shock of gas molecule on solid wall modifies the angle between the speed vector of molecule and the axis of hole. For a molecule entering in hole by the large opening, the angle between the speed vector and the axis of hole increases with a value equal to the opening angle of hole after each shock on solid wall. For a molecule entering in hole by the small opening, the angle between the speed vector and the axis of hole decreases with a value equal to the opening angle of hole after each shock on solid wall. The molecules of gas which enter by the large opening have their trajectories gradually inverted by shocks on wall. These molecules don't reach the small opening of hole. On the contrary, the molecules of gas which enter by the small opening have their trajectories gradually made parallel to the axis of hole by shocks on the oblique wall. These molecules can't return towards the small opening.

The conical hole discriminates the molecules of gas according as they enter by its small or large opening. The molecules of gas entering in hole by the large opening have their trajectories inverted by the shocks on the wall of hole. They quit the hole by the large opening, so their contribution to the pressure applied on the face of the sheet metal is the same as if the holes are absent. On contrary, the molecules of gas which enter in hole by the small opening don't participate on the pressure applied on the face of the sheet metal but, by shocks on the wall of hole, generate a pressure on the opposite face of the sheet metal. If the surface occupied by the small openings represents 1 % of the total surface of the face of sheet metal, it could appear between the two faces of the sheet metal a difference of pressure of 2 %.

Simultaneously at this difference of pressure, a stream of gas through the small openings towards the large openings appears. This orderly movement shows a decrease of disorder of molecules of gas, translated by the decrease of the indicator of this disorder: the temperature. So a difference of pressure could be maintained by the heat of gas in contact. A force, proportional at the area of the sheet metal drilled by conical holes, could be recovered.

If this phenomenon occurs really, it could transform into mechanical form the solar energy stored in the heat of atmosphere. This use of solar energy could have the advantage to be free from the variability of daylight.

One can envisage two modes of manufacture: etching or bombing. Since some years, the single-crystal silicon used in the electronic industry knows new applications for the manufacture of micromechanical devices. The etching is the main tool of this new industry, and more particularly the anisotropic etching. This sort of etching affects the single-crystal silicon in variable speed according as the crystallographic orientation of the silicon surface. It is possible with this technique to make a great variety of holes, particularly conical holes.

The other mode of manufacture could be bombing. One knows how to make metal or ceramic aggregates with a diameter of some nanometers until some hundreds nanometers. Ionized and propelled in high speed by an electrostatic accelerator, these nanometric missiles could generate holes in a sheet metal. If the energy of these solid particles is suitably adjusted, one could, I believe, obtain conical holes. The smallest opening of the hole would be localized in the impact point of the particle and would have its diameter. This mode of ballistic manufacture doesn't necessitate single-crystal silicon as support. Current materials as the steel could be used.

I think that this phenomenon is not incompatible with the second law of thermodynamic. These sheet metal drilled by conical holes could be seen as a dissipative system, permitting the local apparition of negative entropy.

Dau.mic

I haven't finished analyzing it, but this appears to be a simple heat engine. It will cease to function if the Earth and the Sun were in equilibrium. This means that this is no more "direct" conversion than a windmill or a water wheel.