Nature of photoluminescence at a semiconductor heterojunction Do I understand it correctly that photoluminescence at a semiconductor heterojunction occurs because of intralayer recombination? If so, why can't photoluminescence occur because of interlayer recombination? Is it because momentum cannot be conserved in such a process (like in the indirect gap semiconductor?)
 
 A: I'm not too sure what you mean by interlayer and intralayer recombination. 
Photoluminescence occurs in bulk materials without any layers so this isn't an intrinsic requirement.
In the case of LEDs, the forward voltage drives an electron and hole current across the junction. With a high density of electrons and holes in the same volume, the recombination rate becomes large. The electrons and holes recombine and energy is released in the form of photons.
Energy and momentum is conserved in this reaction. 
In direct-bandgap materials, this is easily satisfied because both electrons and holes have very similar momenta (they cluster around the Gamma point). Transitions can be represented as vertical lines in an E-k diagram. 
In indirect materials, the electrons and holes have quite different momenta because the top of the valence band is at a different location to the bottom of the conduction band on the E-k diagram.
This means that optical transition can only occur with the simultaneous absorption or emission of a phonon, which carries little energy but lots of momentum.
Generally, LEDs are made from direct band gap materials. Or they use quantum wells to further confine the electrons and holes in space and increase the emission rate.
Silicon (indirect) LEDs have actually been made! But they required some quite remarkable optical engineering to the surface of the device to reach relatively poor efficiency.
