I thought about a solar furnace prototype. It is like this: light from the Sun is concentrated with a system of parabolic and plane mirrors on the surface of a rock, which is contained inside a vessel. The light and heat is not meant to reach the vessel, only the rock; the vessel could be cooled. The rock can be anything, like granite, basalt or a mineral ore. The light heats the rock to more than 3000 degrees Celsius and the rock materials are effectively vaporised and decomposed into component elements. I.e., the result is a mixture of atomic O, H, C, Al, Fe, Si, Na, K etc., with some compounds still. Now the resulting gases come out of the vessel at high speed and reach a cold surface or cold stream of gases. The problem is: can the metals be separated from lighter elements in this device? I don't intend a perfect (100%) separation, but still a meaningful one (20-50% would probably do). Does it need a flux of reducing gas (say H) to be injected into the vessel in order to avoid reoxidation of metals? Or could it do without it? I think the elements should separate on their own without reoxidation because at a given temperature they have different thermal velocities (heavier ones move slower). The purpose of the device is to be as simple as possible. It should make a good extraction means on other planets/asteroids, but even here on Earth.