What dictates the range of EM radiation which can be produced by semi-conductor excitation (like LEDs) Exciting semiconductors such as silicon carbide and gallium nitride can cause electroluminescence. I believe that by altering the chemistry of the semiconductor, you can alter the wavelength of the radiation emitted? I know that Infra-red LEDs exist and some UV, is there any other non-visible waves which can be emitted and is there a theoretical limit to the wavelength of EM-radiation tht can be produced with this approach?
 A: Here is a pair of bounds:
At the lower energy range, $kT$ is a limiting factor since thermal effects will swamp attempts to invert the electron/hole population.  At room temperature $kT$ is in the infra-red at about 0.025 $eV$.
At the upper energy range you need a conduction band state which is still bound to the crystal.  Among atoms, Helium has the highest ionization energy at about 27 $eV$, which is in the high energy ultraviolet.  That is likely the limit for ionizing any solid.
Thus, semiconductor LED style emission is limited to the IR, the visible, and the UV at room temperature.
I should add that it is not a coincidence that the effect is limited to the visible and to the neighboring energies which are named after "red" and "violet".  When physicists started to classify EM radiation those wavelengths that were associated with molecular transitions but were not visible to the eye were named infrared and ultraviolet.  EM radiation bands not produced by these mechanisms were given different names.
