P-Type Semiconductor

A p-type semiconductor is a semiconductor, usually silicon, doped with acceptor atoms such as boron so that mobile holes outnumber electrons and carry most of the current. The p stands for positive, describing the sign of the hole carriers, not the charge of the crystal, which is electrically neutral overall.

The mechanism is a missing bond rather than a surplus electron. Boron has three valence electrons, one fewer than the silicon atom it replaces, so when it sits on a lattice site one of the four neighboring bonds lacks an electron. A nearby bonding electron can hop into that gap with only a small energy cost, about 0.045 electron-volts for boron in silicon, and doing so completes the bond at the boron atom and leaves a vacancy where the electron came from. On a band diagram the acceptor level lies just above the valence band, and at room temperature, where thermal energy is near 0.026 electron-volts, nearly every boron atom has borrowed an electron from the valence band. The boron atom is now a fixed negative ion, and the vacancy it left behind is free to wander.

Holes are an accounting device that is physically exact: the missing electron moves as though it were a particle with positive charge. When a neighboring bond electron steps into the vacancy, the vacancy appears to step in the opposite direction, and the crystal's collective response matches that of a positive carrier with its own effective mass. In silicon holes move more slowly than electrons, with typical low-doping mobilities near 450 square centimeters per volt-second versus roughly 1,400 for electrons, which is why a PMOS transistor must be drawn wider than an NMOS transistor to deliver the same current. At a boron doping of 10¹⁶ per cubic centimeter the hole density is close to 10¹⁶ per cubic centimeter, while electrons fall to about 10⁴, because the product of the two densities stays fixed near the square of the intrinsic density.

The Hall effect is the classic way to tell p from n, because the sign of the carrier shows up as the sign of a voltage. Edwin Hall found in 1879 that a current in a conductor placed in a perpendicular magnetic field produces a transverse voltage; for electrons the polarity is one way and for holes it is the opposite. Measuring that voltage in a bar of semiconductor gives the Hall coefficient, whose sign identifies the dominant carrier and whose magnitude gives the carrier density. The labels n-type and p-type came out of the Bell Labs silicon work of around 1940, where the two kinds of material were distinguished by measurable electrical behavior, and Hall measurements remain a standard characterization tool for wafers and thin films today.

The persistent misconception is that p-type silicon is positively charged overall. The material is exactly as neutral as n-type silicon, because each mobile hole is balanced by a fixed negative boron ion locked in the lattice. What is positive is the carrier, not the block. Boron-doped p-type wafers are the most common substrate for mainstream CMOS chips, with n-type wells carved into them for the p-channel transistors, and the p-type regions in every pn junction, solar cell and bipolar transistor owe their behavior to the same trick. Germanium devices of the 1950s reached p-type with gallium or indium instead, and gallium-doped silicon is now used in some photovoltaics to avoid a boron-oxygen defect that slowly cuts cell efficiency under light.

Read p-type semiconductor as silicon with deliberately missing bond electrons, in which a neutral crystal lets positive holes drift while fixed negative ions stay put.