Fermi level makes doped and undoped different?

[Outrageous]

The intrinsic semiconductor
Fermi level is the maximum energy level an electron can be in at 0K.
Or Fermi level is a level where there is a probability of 50% to find electrons at any temperature.
Conduction band is a range of energy where electrons freed from bonding stay.
Valence band is a range of energy where electrons in a bonding between atoms stay.
In undoped semiconductor, the Fermi level is in between the energy gap, which is lower than conduction band but higher than the valence band.
We studied that when an electron absorbs thermal energy, it will be excited to the conduction band by crossing the band gap which is where the Fermi level at.(Why do we need to define Fermi level? It is used for?)
Is that because...
In extrinsic semiconductor, the Fermi level of n-type shift to the conduction band, while that of p-type shift to the valence band. ( Is that what making the doped semiconductor having higher conductivity? So that is also the reason affect the density of hole and electron? Then how can the conductivity of doped semiconductor increase just by shifting the Fermi level? Let’s say n-type, the Fermi level shift towards the conduction band but the conduction band is still stay at the same energy level, then that means electrons in valence band still have to absorb same amount of energy as in undoped semiconductor in order to excite to the conduction band. So the only thing that makes doped and undoped different is the Fermi level and the density of eletron. But how is that going to affect? Apart from that, my book says the n-type has higher conductivity because there are electrons from pentavalent atoms which are loosely bonded, so how should I relate the Fermi level of n-type with the loosely bonded electrons? )


Please help to check whether my concept is correct and answer me. Thank you.

 

[mzh]

You're partially confused. The Fermi level is the energy of a hypothetical state with probability of being occupied equal 50%.
To realize this, you need to consider the Fermi distribution function f(E) which tells you how probable it is that a state is occupied. You will realize that under equilibrium conditions, if a state is located at an energy larger than E_F + 3*kT, it is highly unprobably that it is occupied. Conversely, if it is located at E_F - 3*kT, it is extremely probably that it will be occupied.

Then how can the conductivity of doped semiconductor increase just by shifting the Fermi level?

The Fermi level shifts *because* of the doping, not the other way around.