Electron Hole

An electron hole is the absence of an electron in a nearly full band, typically the valence band of a semiconductor, which behaves in every electrical respect like a mobile particle with positive charge. When a bond in a silicon crystal is broken and an electron leaves, the empty spot it leaves behind can be filled by a neighboring bond electron, which in turn leaves its own empty spot; tracked from outside, the vacancy appears to travel through the crystal in the direction of the electric field, as a positive charge would. This bookkeeping device turns the messy motion of around 10^23 valence electrons per cubic centimetre into the simple motion of a handful of positive carriers.

The hole concept came from trying to explain a strange experimental fact: some materials showed a positive sign in the Hall effect. Edwin Hall had discovered in 1879 that a magnetic field deflects current to produce a sideways voltage whose sign reveals the charge of the carriers. Several metals and semiconductors gave a sign corresponding to positive carriers, which no free-electron theory could explain. Rudolf Peierls in 1929 and Werner Heisenberg in 1931 showed that in a band that is almost filled, the missing electrons respond to fields as positive particles would. That is the origin of the idea, and it should not be confused with Paul Dirac's hole theory of 1928-30 for antimatter, where a vacancy in a sea of negative-energy states is interpreted as the positron; a semiconductor hole is a real many-electron effect in a solid, not a new elementary particle.

A hole travels because neighbors move, not because an atom does a row of five bond sites; an electric field pushes electrons to the left Step 1 hole at site 2 Step 2 hole at site 3 electron moves left Step 3 hole at site 4 hole drifts right, carries positive charge

The hole is useful because it simplifies the physics rather than because anything positive physically moves. In a nearly full band, summing the motion of every electron is intractable, but the total current equals that of the few empty states acting as particles with charge +e. The key correction is effective mass: the crystal's periodic forces make a hole respond to a push as though it had a particular mass, which in silicon is about 0.37 of the free-electron mass for conduction purposes, versus about 0.26 for electrons. William Shockley's 1950 book Electrons and Holes in Semiconductors set out the theory, and the Haynes-Shockley experiment of 1949 at Bell Labs, in which James Haynes and Shockley watched a pulse of injected carriers drift and spread in germanium, showed holes moving with measurable mobility. That experiment, and the point-contact transistor before it, rested on injecting these carriers.

Photon or heat frees one bond electron
needs at least the band gap, about 1.12 electron-volts in silicon
An electron-hole pair now exists
Electron in the conduction band
mobility around 1400 cm^2 per volt-second in lightly doped silicon
Hole in the valence band
mobility around 450 cm^2 per volt-second
Field applied: carriers drift in opposite directions
both contribute current in the same direction in the circuit
Recombination: electron falls back into the hole
releases heat or, in direct-gap materials, a photon

Holes are slower than electrons in silicon, and the difference shapes how chips are laid out. Mobility, the ratio of drift velocity to electric field, is about 1400 square centimetres per volt-second for electrons in lightly doped silicon at room temperature and about 450 for holes, a ratio of roughly three; in germanium the values are higher, about 3900 and 1900. Holes are slower because valence-band states are more strongly warped and scattered between sub-bands and because their effective mass is larger. In complementary logic the p-type transistors, which use holes, were therefore drawn wider than the n-type ones to equalize drive current, and strained silicon-germanium channels, introduced in production around 2003, were developed largely to boost hole mobility.

The central pitfall is to picture a hole as a tiny particle or an atom hopping through the lattice. The atoms stay in place; what moves is the vacancy, through successive jumps of bonded electrons into it, and the hole has a positive effective charge only as a description of the collective response of the filled band. Equally common is the mistake of thinking that a hole is simply a missing atom: that is a vacancy, a lattice defect, while a hole is a missing electron. Hole conduction is real, though, in every p-type region, in the base of a p-n-p transistor, and in the channel of every p-type MOSFET inside a processor or AI accelerator.

Read an electron hole as the empty seat in a nearly full theatre: no one seat travels, yet as neighbors shuffle into the gap the vacancy moves the other way, and that bookkeeping lets physicists and engineers treat a crowd of electrons as a few positive carriers.

Take electron hole further

Ask the copilot about this term, or have our engineers assess it against your process.