The PN junction

Circuit Analysis was built entirely from elements that treat both directions alike: a resistor does not care which way current flows, and neither does a capacitor or an inductor. Every element you have met is symmetric and linear. Microelectronics begins by breaking both properties at once, with a device made of two kinds of doped silicon pressed together. The result conducts one way and blocks the other — and that asymmetry is the seed of rectifiers, logic, memory, and every transistor in the machine you are reading this on.

Pure silicon is a poor conductor: each atom's four valence electrons are locked into covalent bonds. Doping deliberately contaminates it. Add a pentavalent atom such as phosphorus and its fifth electron is left over, free to roam: the result is n-type silicon, where the mobile charge carriers are electrons. Add a trivalent atom such as boron and a bond is left short one electron — a hole — which neighbouring electrons hop into, so the vacancy itself migrates like a positive particle: p-type silicon, where the carriers are holes. Both materials are electrically neutral overall; doping changes what can MOVE, not the net charge.

Now join them. Free electrons near the boundary diffuse into the p-side and fall into holes; holes diffuse the other way. Each crossing leaves behind a fixed, charged dopant ion — negative on the p-side, positive on the n-side — and this exposed layer of immobile charge is the depletion region, so called because the mobile carriers there have been consumed. The charge it holds creates an electric field pointing from the n-side to the p-side, which opposes further diffusion. Equilibrium arrives when the field's opposition exactly balances the diffusion tendency, leaving a built-in potential across the junction: roughly 0.7 V for silicon, about 0.3 V for germanium.