Loading field map
Semiconductor materials are substances whose electrical conductivity sits between that of conductors, like copper, and insulators, like glass or rubber. This intermediate behavior is not merely a weaker version of metallic conduction; it arises from a specific electronic structure that allows their conductivity to be finely and reversibly controlled. This tunability—through chemical doping, applied electric fields, light, or temperature—is the foundation of virtually all modern electronics, from microprocessors and memory chips to light-emitting diodes (LEDs), lasers, and solar cells. The field of semiconductor materials is the study of this class of substances: what makes them semiconducting, how their properties can be engineered, and how they can be synthesized into the highly perfect crystals required for devices.
The defining feature of a semiconductor is its band gap: an energy range that electrons cannot occupy. In a conductor, the highest occupied electron band and the lowest empty band overlap, so electrons can move freely. In an insulator, the gap is so large that thermal energy at room temperature cannot promote electrons across it. In a semiconductor, the gap is small enough (typically between about 0.1 and 3.5 electron-volts) that electrons can be excited across it by heat, light, or an applied voltage, but large enough that the material's conductivity can be switched on and off.
The central practical question is how to control the number and movement of charge carriers—electrons in the conduction band and the "holes" (missing electrons) they leave behind in the valence band. This control is achieved through two primary means. The first is doping: the deliberate introduction of impurity atoms. A silicon atom has four valence electrons; replacing one with a phosphorus atom, which has five, donates an extra electron, creating an n-type (negative) semiconductor. Replacing it with a boron atom, which has three, creates a missing bond, or hole, which behaves as a positive charge carrier, making a p-type semiconductor. The second means is the electric field effect, where an external voltage attracts or repels carriers near a surface, forming a conductive channel that can be turned on and off. The interplay of these two controls, especially at junctions between n-type and p-type regions, produces the rectifying behavior of diodes, the amplification of transistors, and the light emission of LEDs.
The history of semiconductor materials is not a single linear story but a succession of material systems, each solving different problems. The earliest practical semiconductor devices, in the early twentieth century, used selenium and copper oxide for rectifiers in power applications. These were polycrystalline materials with limited performance, but they established that certain non-metals could control current flow.
The first truly transformative material was germanium. In the late 1940s, the invention of the point-contact and then the junction transistor at Bell Labs used germanium because it could be purified to an extraordinary degree and doped controllably. Germanium's small band gap (0.66 eV) made it easy to work with, but it also meant that devices leaked current at elevated temperatures, limiting their practical use.
Silicon replaced germanium in the late 1950s and 1960s for several decisive reasons. Its larger band gap (1.12 eV) greatly reduced thermal leakage. More importantly, silicon forms a native oxide, silicon dioxide (SiO₂), which is an excellent electrical insulator and can be grown thermally on the silicon surface. This oxide serves as a mask for selective doping and as the gate dielectric in the metal-oxide-semiconductor field-effect transistor (MOSFET), the structure that became the workhorse of digital electronics. No other semiconductor has a native oxide with comparable properties, which is why silicon dominates the microelectronics industry to this day. The development of the integrated circuit—fabricating many transistors on a single silicon wafer—turned this material advantage into an economic and technological revolution, following a scaling trend (often called Moore's law) that has held for decades.
While silicon is ideal for logic and memory, its indirect band gap means it is a poor emitter of light. When an electron falls across the gap, the energy is released mostly as heat (phonons) rather than photons. This limitation drove the development of compound semiconductors, which are made of two or more elements.
The most important family is the III-V semiconductors, named for their positions in the periodic table: compounds of group III elements (aluminum, gallium, indium) with group V elements (phosphorus, arsenic, antimony). The archetype is gallium arsenide (GaAs), which has a direct band gap, meaning electrons can recombine with holes to emit light efficiently. This property enabled the first practical semiconductor lasers and LEDs in the 1960s and 1970s. GaAs also has higher electron mobility than silicon, making it useful for high-frequency communications.
The key advantage of compound semiconductors is band gap engineering. By forming alloys—such as aluminum gallium arsenide (AlGaAs) or indium gallium arsenide (InGaAs)—one can continuously tune the band gap and the lattice constant across a range. This allows the construction of heterostructures: layered crystals where the band gap changes abruptly at interfaces. Heterostructures confine electrons and holes to very thin layers, dramatically improving laser efficiency and enabling the high-electron-mobility transistor (HEMT), which is used in radar and satellite communications. The same principle underlies quantum wells, where carriers are confined in one dimension, and more exotic structures like quantum wires and dots.
A second major family is the II-VI semiconductors, such as zinc selenide (ZnSe) and cadmium telluride (CdTe). These cover a wider range of band gaps, including the visible spectrum, and are used in blue-green LEDs (though largely superseded by III-nitrides) and in thin-film solar cells. Lead chalcogenides (lead sulfide, lead selenide) are narrow-gap materials used in infrared detectors.
A distinct and increasingly important class is the wide-band-gap semiconductors, with gaps above about 2.5 eV. These include gallium nitride (GaN) and silicon carbide (SiC), as well as the compound aluminum nitride (AlN) and the element diamond. Their large gaps allow them to sustain very high electric fields without breaking down, making them ideal for high-power switching in electric vehicles, power grids, and radio-frequency amplifiers. GaN also emits blue and ultraviolet light; the development of efficient blue LEDs and lasers based on GaN in the 1990s completed the RGB color palette for solid-state lighting and enabled Blu-ray discs. SiC has the additional advantage of a native oxide (SiO₂) that can be grown on its surface, allowing for MOSFET-like devices, though the interface quality is inferior to that of silicon.
These materials are harder to grow and process than silicon or GaAs. They require very high growth temperatures, and their substrates are expensive and often defective. Yet their performance advantages in high-power and high-frequency applications are so significant that they have carved out a major commercial niche, coexisting with rather than replacing silicon.
A more recent and conceptually distinct approach is the use of organic semiconductors: carbon-based molecules and polymers with conjugated π-electron systems. These materials can be processed from solution at low temperatures, enabling flexible, lightweight, and low-cost electronic devices such as organic LEDs (OLEDs) for displays, organic solar cells, and simple sensors. Their charge-carrier mobilities are far lower than those of crystalline inorganic materials, and they are more susceptible to degradation, so they do not compete with silicon for high-performance logic. Their value lies in applications where mechanical flexibility, large-area coverage, and low-cost manufacturing matter more than raw speed.
A related emerging area is perovskite semiconductors, named after the crystal structure of the mineral calcium titanate. Hybrid organic-inorganic lead halide perovskites have shown remarkable efficiency gains in solar cells within a short period, rivaling silicon in laboratory settings. However, they face serious challenges of stability and toxicity, and their long-term commercial viability remains an open question.
The extraordinary performance of semiconductor devices depends on the ability to grow crystals of exceptional purity and structural perfection. The dominant method for bulk crystals is the Czochralski process, in which a seed crystal is dipped into a melt of the semiconductor and slowly withdrawn, pulling a large single crystal ingot. For silicon, this produces ingots up to 300 mm in diameter. For compound semiconductors, which decompose at high temperatures, the Bridgman method or liquid-encapsulated Czochralski is often used, where the melt is covered by a molten glass layer to prevent volatile component loss.
For device structures, the critical technique is epitaxy: the growth of thin, single-crystal layers on a crystalline substrate. Molecular beam epitaxy (MBE) and metal-organic chemical vapor deposition (MOCVD) are the two dominant methods. MBE uses beams of atoms or molecules in an ultra-high vacuum, allowing atomically precise control of layer thickness and composition. MOCVD uses chemical precursors carried by a gas stream, which react on the heated substrate; it is faster and more scalable, making it the standard for commercial production of LEDs and lasers. Both methods enable the fabrication of heterostructures with interfaces that are abrupt to within a single atomic layer.
The field of semiconductor materials today is not unified by a single theory or school but by a shared set of questions: how to control band structure, how to manage defects, and how to integrate different materials into functional devices. The dominant material remains silicon, whose processing infrastructure is unmatched. But the field is increasingly characterized by heterogeneous integration: combining silicon logic with III-V lasers, GaN power amplifiers, and other specialized materials on a single chip or package. This is driven by the physical limits of silicon scaling and by the demand for new functionalities, such as optical communication and high-voltage power conversion.
A central tension is between performance and manufacturability. Silicon wins on cost and integration but loses on optical and high-frequency properties. GaN and SiC win on power handling but are expensive to grow and process. Organic and perovskite materials win on flexibility and low cost but lose on stability and speed. The field's progress is not a matter of one material defeating another but of matching each material's strengths to specific applications, and of developing new techniques—such as wafer bonding, selective area growth, and two-dimensional materials like graphene and transition-metal dichalcogenides—to bridge the gaps between them.
Another enduring tension is between purity and functionality. The highest-performance devices require near-perfect crystals with controlled impurity profiles, yet the most interesting physics often emerges from deliberately engineered disorder, such as doping, alloying, or strain. The field's history is a series of solutions to this tension: silicon's native oxide solved the surface-state problem that plagued germanium; heterostructures solved the carrier-confinement problem in lasers; and wide-band-gap materials solved the breakdown problem in power devices. Each solution has opened new questions, and the field remains defined less by a settled doctrine than by the ongoing effort to understand and control the electronic behavior of crystalline matter.