Can Doping Transform Any Isolator into a Semiconductor?

[Quasi Particle]

Semiconductors and isolators are only different in that the energy gap of the former is smaller than 2eV, and of the latter it is larger. I can change the energy gap by doping - but can I make any isolator a semiconducter? What are the limitations?

 

[Barth]

You can not really "change the gap" by doping a cristal. When you dope the cristal, you introduce impurities in the lattice. These impurities creates localised states for the electrons, whose energy is inside the gap. So you don't change the gap, you only create disrete levels inside.

All author's don't really agree for the difference semiconductors/isolators. When you cite a limit a 2eV, it is a little bit arbitrary.
The difference is that in a isolator, the energy gap is many ordrer of magnifude above the thermal energy kT at room temperature ( 25meV) , so the thermal excitation of electron and hole is negligible. In other words, there is no charge carriers at room temperature.
Instead, an element whose gap Eg is smaller than -let's say- 100*kT ( which for the room temperature of 300 K, means Eg<2 eV), presents a non-negligible occupation of the conduction band states, and then behaves as a semiconductor.

In summary, if you want that your isolator becomes a semiconductor, you just have to put it in a oven.

And for the doping limitations, i can't tell you. If i understood well, there is different effects that can happen when you dope too much an element. The first is that in highly doped elements, the fermi level comes into the conduction band. Your element then begin to have a metal-like behaviour.
The others are more technoligic: the doping of a material is not a trivial operation. Doping by ionic implantation, for instance, can heavily damage the sample if the concentration of dopants is too high.