I need someone familiar with supercolliders, those things they use to smash matter and antimatter together, to help me explore this. On its "antimatter" page, Wikipedia cites a source (http://www.nasa.gov/exploration/home/antimatter_spaceship.html) claiming that antimatter, in the form of positrons, can be produced at $250million per 10 milligrams. 10 mg is a lot of antimatter! If that were reacted with another 10 mg of normal matter, it would create 1,800 gigajoules of energy! Assuming we could harness that energy into electric power, and assuming we could throttle the collision rate, then we could have for example 1 megawatt of power for almost 21 days. Of course it's not going to be 100% efficient, but even a system efficiency of 25% would still be pretty darn good imo. Here's my concept. Electrons and positrons are held in separate chambers of a device, probably by a magnetic field. Somehow they are made to move out of those chambers at a very small rate, although probably at high speeds, and collide with each other in a collision chamber. This produces gamma rays. The collision chamber walls are lined with solar panels, tuned to absorb gamma rays, and via the photoelectric effect convert it to DC electric power. Some of this power would go back into maintaining the magnetic containment fields so that the device is self-sustaining. To me it seems great: no moving parts and better energy density than nuclear power. Too good to be true? Can anyone vouch for this or explain why it's impractical? I can think of several engineering problems and hopefully we can address most of them. (1) Is it possible to create solar panels that are NOT tuned to absorb visible or infrared light, as normally is the case, but rather gamma rays? Also they would have to have a decent lifetime. (2) How to contain the separate blobs of electrons and positrons in a safe way, yet still have a way to accelerate them out of the chamber so they collide with one another at a small rate? We can accelerate one or the other, or maybe both. (3) Do we even need to contain electrons in their own chamber? Or can we just liberate electrons on demand from normal conducting material? (4) Would the solar panels have to be cooled? (5) Exactly how much electric power is needed to magnetically contain a blob of X grams of positrons? (in a vacuum?) And exactly how much electric power is needed to accelerate them to the collision chamber? (6) Can we ensure that a high percentage of the matter and antimatter, say over 90%, indeed collide within a targeted volume and don't miss? (7) Is there a better choice of matter-antimatter? Ignoring procurement problems, would it be better to use protons and antiprotons? Possibly hydrogen and antihydrogen or helium and antihelium? An isotope of them? (8) Instead of solar panels, is there a better way to convert gamma rays into electric power? Note, im NOT thinking of this as a cheap or renewable energy source of the future. $250M per 1,800 GJ comes out to $500 per killowatt-hour! Im thinking of this as a great power source for spacecraft, maybe submarines as well, or any remote vehicle.