N-doped region tend to flow to the p-doped region

[Harmony]

When a diode is first manufactured, the electrons in the n-doped region tend to flow to the p-doped region due to diffusion effect. In the process, a depletion region is formed until the electric field in the depletion region cancel out the diffusion effect. Hence an equilibrium is reached. (or so I thought)

However, since this two effect cancel out each other exactly, why would we need a minimum voltage of about 0.3-0.5 V to forward bias it? Wouldn't any voltage cause current to flow, since the diffusion+external voltage source is now greater than the electric field in the depletion zone alone?

 

[MATLABdude]

You're correct in that a depletion region is formed, and that a "built-in" potential across the depletion region results, but incorrect in the ramifications of this. The built-in potential opposes the direction of current flow as it normally occurs when the diode is acting in forward-bias mode.

So you need to overcome the built-in potential (the forward diode voltage) before the diode turns on and current starts flowing:
http://en.wikipedia.org/wiki/Diode#Current.E2.80.93voltage_characteristic