Defects and Impurities in Semiconductors

Abstract

Impurities are essential for giving semiconductors the properties that render them useful for electronic and optoelectronic devices. The intrinsic carrier concentrations in most semiconductors are quite low. Adding small amounts of impurities allows control of the conductivity of the semiconductor: shallow donors, such as phosphorous in silicon, produce n-type conductivity (carried by electrons), and shallow acceptors, such as boron in silicon, produce p-type conductivity (carried by holes). These doped layers and the junctions between them control carrier confinement, carrier flow, and ultimately the device characteristics. Commonly used semiconductors such as Si and GaAs can be doped both p-type and n-type. Constraints on doping still limit device performance, however. For instance, the shrinking size of Si field-effect transistors requires higher doping densities, with donors exhibiting deactivation when the doping increases above ∼3 × 1020 cm−3. Doping problems have been more severe in wide-band-gap semiconductors such as ZnSe, GaN or ZnO, which typically exhibit unintentional n-type conductivity, and in which p-type conductivity has been difficult to achieve. Point defects (vacancies, self-interstitials, and antisites) have often been invoked to explain these difficulties.