Photoelectric effect Why the photoelectric effect is observed only for metals and not for non-metals? 
Isn't it possible for the photon to release an electron from a non-metal surface?
 A: Conduction electrons in metals lie in the conduction band, they have a very low binding energy compared to electrons in an insulator, for instance. It thus requires a low amount of energy to excite them and have a photoelectric effect.
You can see on this article showing original data from Milikan, that the minimum amount of energy per photon required to have a photoelectric effect with sodium is of order of $\Delta E_{\text{conductor}} \simeq 1,8$ eV.
In insulators, the highest-energy electrons lie in the valence band and you need a huge amount of energy to excite them : it is harder to have a photoelectric effect in this case. In the case of insulators, this energy is of order of magnitude of the first ionization energy of the atom or molecule which the solid is made of.
You can see here a table giving in eV the first ionization energy of each atom. (It's in french, couldn't find the same table on English wikipedia). You see that the energy required is of order of $\Delta E_{\text{insulator}} \simeq 10$ eV for most insulating pure compounds. It is thus possible to have a photoelectric effect, but would require photons with ten times more energy than for insulators.
A: The outer electrons in insulators tend to have higher binding energies, so it requires more photon energy to free the electron. So, the photoelectric effect is still possible in insulators, it's just harder.
