In a semiconductor, how exactly do electrons move into the conduction band? From what I know, electrons that are excited can move into the conduction band, but how and what causes electrons to be excited? Heat/light seem to make sense as to something that can 'excite' the electron, but I am not completely sure.
Does the location of the electron matter at all, or can an electron at the lowest energy of the valence band be excited to the conduction band (even if there is an acceptor defect above it, could it possibly skip that and move straight to conduction)?
 A: There are many processes that may excite electrons into the conduction band:

*

*light (even if we do not illuminate the semicodnuctor, there is usually some background radiation)

*strong electric field (Zener tunneling)

*collisions with other electrons, in which they may exchange energy

*scattering particles, such as neutrons against the semiconductor

and others.
If nothing special is done in terms of forcing excitations, it is usually background radiation that causes the excitation. In statistical mechanics terms the semiconductor can be thought of as a subsystem in thermal equilibrium with its surroundings, which enables us to use the Boltzmann/Fermi distribution without thinking of the exact processes that cause the thermal equilibrium (which are usually neglected in statistical physics, but whose existence is always implied).
Let me note that heat is not the same as light, but rather a general measure of energy exchange. In this case it is indeed light (meaning generally electromagnetic radiation, not necessarily visible spectrum).
The probability of transition from this or that place in the valence band is governed by the energy conservation and the selection rules (i.e., the magnitude of the matrix element for the specific type of excitation mechanism).
