A semiconductor is a material whose electrical conductivity can be controlled by impurities, temperature, illumination, or electric fields. Its conductivity commonly lies between that of a metal and an electrical insulator, although this comparison alone does not fully define the category. Semiconductor materials provide the basis for transistors, integrated circuits, and many optical and power-conversion devices. The term also informally denotes components manufactured from these materials, rather than the materials themselves. Their behavior is explained through quantum mechanics and the electronic structure of solids. (ocw.mit.edu)
Electronic structure and charge carriers
In a solid, interactions among atoms produce ranges of allowed electron energies called bands. In a conventional semiconductor, the valence band and conduction band are separated by a band gap: an energy interval containing no allowed states in the ideal bulk crystal. An electron can enter the conduction band after receiving sufficient energy through thermal excitation or absorption of light. Unlike a metal, an ideal, undoped semiconductor has a filled valence band and an empty conduction band at absolute zero. (ocw.mit.edu)
Excitation leaves an unoccupied state in the valence band, called an electron hole. As neighboring electrons fill this vacancy, the vacancy behaves as a mobile carrier of positive charge. Both conduction-band electrons and holes contribute to electrical current. In an intrinsic semiconductor—one whose carrier population is not dominated by impurities—electron and hole concentrations are equal. Heating generally increases the intrinsic carrier concentration, while the resulting conductivity also depends on how readily those carriers move. (ocw.mit.edu)
For ordinary low-field transport, conductivity is expressed as
[ \sigma=q(n\mu_n+p\mu_p), ]
where (q) is the magnitude of the elementary charge, (n) and (p) are electron and hole concentrations, and (\mu_n) and (\mu_p) are their mobilities. This distinction between carrier concentration and mobility is important: adding carriers does not necessarily improve every aspect of device performance. (ocw.mit.edu)
Doping and junctions
Doping deliberately introduces impurities to alter carrier populations. In silicon, phosphorus commonly acts as a donor, supplying electrons and producing n-type material. Boron acts as an acceptor, increasing the hole population and producing p-type material. The letters indicate the predominant mobile carrier, not the net electrical charge of the entire crystal; doped material remains approximately neutral away from charged interfaces. Dopants introduce energy levels that require much less excitation energy than the full band gap. (ocw.mit.edu)
When p-type and n-type regions meet, they form a p–n junction. Carrier diffusion across the interface leaves behind ionized dopants, creating a depletion region and an internal electric field. At equilibrium, this field opposes further net diffusion. An applied voltage changes the barrier: forward bias favors carrier injection, whereas reverse bias generally suppresses current until leakage or breakdown mechanisms become significant. This asymmetric response underlies the operation of a junction diode. (ocw.mit.edu)
A metal–oxide–semiconductor field-effect transistor, or MOSFET, uses a voltage applied to an insulated gate to control a conducting channel between source and drain. Unlike a simple junction diode, it permits one electrical signal to regulate another current, enabling amplification and switching. (live.ocw.mit.edu)
Materials and optical properties
Silicon and germanium are elemental semiconductor materials; gallium arsenide and gallium nitride are compound semiconductors. Material selection depends on electronic structure, carrier transport, processing requirements, and the intended operating conditions. Semiconductor technology therefore encompasses more than silicon computer chips. (ocw.mit.edu)
An important optical distinction is whether the band gap is direct or indirect. In a direct-gap material, the conduction-band minimum and valence-band maximum occur at the same crystal momentum, favoring radiative transitions. In an indirect-gap material, momentum must also change, commonly through interaction with a phonon. Ordinary crystalline silicon has an indirect gap, while gallium arsenide and gallium nitride have direct gaps. These differences help determine suitability for light emission and absorption. (arxiv.org)
Wide-bandgap semiconductors, including silicon carbide and gallium nitride, support devices designed for high voltages, high switching frequencies, or elevated temperatures. Their properties can reduce losses in power conversion, but practical performance also depends on device structure, thermal management, and packaging. They are not interchangeable replacements for silicon in every application. (energy.gov)
Devices and applications
Semiconductor devices perform several distinct functions. Transistors provide switching and amplification; integrated circuits combine interconnected components to implement processing, memory, and control functions. These circuits underpin computers and other electronic systems. Power semiconductor devices regulate and convert electrical power, including in motor drives and renewable-energy equipment. (educationalgames.nobelprize.org)
Optoelectronic devices connect electrical and optical processes. In a solar cell, absorbed light generates carriers that a device structure separates and collects. In a light-emitting diode, electron–hole recombination produces light; semiconductor lasers additionally use stimulated emission. Band structure, junction design, and material composition determine the wavelengths and efficiencies available. (ocw.mit.edu)
Manufacturing and historical development
Chip fabrication builds patterned structures on a semiconductor wafer through repeated processing steps. Thin-film deposition adds materials, photolithography defines patterns in a light-sensitive coating, and etching removes selected regions. Ion implantation introduces dopants. Repeated patterning and processing create devices and their connections; finished chips are separated and packaged for electrical connection and protection. Lithography defines patterns rather than directly completing all the structures it depicts. (asml.com)
The first working point-contact transistor was developed in 1947 by John Bardeen and Walter Brattain at Bell Labs. They and William Shockley received the 1956 Nobel Prize in Physics for semiconductor research and the discovery of the transistor effect. Jack Kilby and Robert Noyce developed integrated-circuit approaches in 1958 and 1959, respectively, enabling multiple components to be manufactured together on semiconductor material rather than assembled individually. (educationalgames.nobelprize.org)