Edit: instead of using neutrons emitted from a rod, could it instead use alpha particles emitted from a sheet, the radioisotope perhaps being the commonly used plutonium 238. This sheet either resembling solar sails. Or maybe being a rose like shape open at the back, this rose shape would probably be much less propulsively efficient because alpha particles would likely undergo many reflections before being ejected, but might make up for that by potentially having much greater surface area and perhaps being much lighter for a given amount of surface area The sheet would consist of a main layer material that would reflect the alpha particles. Alpha particles should be very easy to reflect. Perhaps this could be a thin aluminum foil (or is there a more suitable material?). On that layer would be a thin layer of the alpha emitter (pu238 maybe?) that would be deposited onto the reflector layer. The α emitting layer would probably need to be incredibly thin as to not absorb the alpha particles that it produces. Excess heat would probably just be given off radiatively. This means the probe would need a separate energy source. Electromagnets might be used to deflect the ejected α particles to produce torque on the body, or the sail could partially retract on one side to change the center of thrust. This would be done to desaturate the reaction wheels of the reaction control system. The propulsion cannot be throttled, but if need be it can simply periodically rotate 180 degrees to produce net zero thrust over time. These sails would be used for long term mission and produce only a very small amount of thrush.
Old but somewhat relevant section: Is it possible for a propulsion system that uses neutrons directly for space propulsion. The high speed particles are produced by a radioisotope fuel rod, it would likely be long and very narrow to decrease the amount of neutron kinetic energy absorbed. Neutrons ejected backwards would produce an impulse for the craft. Neutrons that move forward would be reflected by a neutron mirror imparting a forward impulse on the craft. This neutron reflectors's shape is not greatly considered in this hypothetical. The radioisotope fuel rod would be cooled using very small channels on the inside being cooled by a circulating fluid. This system would combine the propulsion and a radioisotope generator into one. The direction of thrust relative to the the craft's center of mass fan be done by articulating the entire thruster. It could not, however, modulate the thrust output of the thruster. My idea is that it wont use any additional propellant, the only propellant used would be the high velocity neutrons emitted from the radioisotope. Would this work? Could it rival a conventional radiothermal powered electric ion rocket that carries additional propellant?
