doping of semiconductors

Pure silicon is a rather poor conductor — only a trickle of electrons can cross its band gap at room temperature. Yet silicon runs the entire digital world. The trick that transforms it is doping: deliberately stirring in a tiny, controlled amount of a different element, often just one foreign atom in a million silicon atoms. That whisper of impurity changes the electrical behaviour by orders of magnitude, and it is the single most important idea in all of solid-state electronics.

Silicon has four valence electrons and bonds to four neighbours, using up every electron. Now replace one silicon atom with phosphorus, which has five valence electrons. Four of them make the normal bonds; the fifth is left over, loosely held, and easily promoted into the conduction band where it roams free. This is n-type doping (n for negative): the dopant donates extra electrons, the majority carriers. Alternatively, replace a silicon with boron, which has only three valence electrons. One bond is left short an electron — a hole, an empty slot in the valence band. Neighbouring electrons can shuffle into the hole, which makes the hole itself appear to move like a positive charge carrier. This is p-type doping (p for positive). In both cases the dopant creates new energy levels just inside the band gap, so far less energy is needed to make charge carriers than in the pure crystal.

Doping matters because junctions between n-type and p-type silicon are the building blocks of everything electronic. Put a p-region next to an n-region and you have a diode that lets current flow one way only; stack them as n-p-n or p-n-p and you have a transistor, the switch and amplifier that, replicated billions of times on a chip, is the basis of every computer and phone. Add light and a p-n junction becomes a solar cell or an LED. So the abstract band picture of solids, plus the deliberate introduction of a few defects, is the literal foundation of the information age.