The Field That Pushes Backward
The Field That Pushes Backward
Push a charge carrier through silicon hard enough and its drift velocity climbs toward roughly 1×10^7 cm/s, then stops climbing. Push harder still and nothing more happens: velocity holds flat near that ceiling no matter how much stronger the field becomes, because the carrier is now spending its added energy on phonons and photons instead of speed. Ohm's law quietly stops applying. A current limiter can be built out of this fact alone — a device that passes the same current across a wide range of voltage, simply because there is nowhere left for the current to go.
Push a charge carrier through gallium arsenide or indium phosphide the same way and the story does not end at a ceiling. Velocity rises toward a peak just as in silicon — then, as the field keeps increasing past that peak, velocity actually falls. The carrier is not losing energy to heat; it is being kicked into a different conduction band, one whose electrons simply move slower at the same field strength. More push produces less speed. Current in these materials can decrease as voltage rises further, a negative differential resistance with no analog in the silicon case. This one material fact is the entire operating principle of the Gunn diode: not a rectifier, not a switch, but a two-terminal device built on the fact that its own carriers travel backward from where more field would suggest.
Same cause — an electric field pushed past the point where an ordinary conductor would keep obeying it. Two different endings, decided entirely by which conduction bands the material happens to have waiting above the one the carrier started in. Silicon has nowhere lower to fall, so it plateaus. Gallium arsenide has somewhere slower to fall into, so it does — and a whole device family exists because of that one difference in the band structure, not because anyone designed it there on purpose.