Nonstoichiometric Defects in Silicon-Doped GaAs Epilayers Grown on (lll)A- and (lll)B-Oriented Substrates

摘要

Epitaxially grown gallium arsenide (GaAs) has retained for many years its position among the main semiconductor materials, in particular, for optoelectronic devices and modern nanoelectronics. In growing GaAs by molecular beam epitaxy (MBE) on (100)A-and (lll)B-oriented substrates, epitaxial layers (epilayers) with n-type conductivity are traditionally obtained using silicon as the dopant, which builds-in predominantly into the gallium sublattice, behaves as a stable donor impurity, and provides a controlled free electron concentration of up to 6 x 1018 cm"3. For Si-doped GaAs epilayers grown on (Xll)A (X = 1, 2, 3) substrates, both the conductivity type and the free electron concentration depend on the ratio of partial pressures of arsenic and gallium during the MBE process [1-4]. At a low partial pressure of arsenlg, the main fraction of silicon atoms occupies sites in the* arsenic sublattice (SiAs), which leads to thep-type conductivity, while an increase in the arsenic pressure is accompanied by an increasing fraction of Si atoms occupying gallium sites (SiGa), which results in a change of the conductivity type. This property of silicon can be useful for solving some technological problems encountered in the development of GaAs-based devices. Indeed, using the same dopant and substrates containing regions with different orientations on the growth surface, it is possible to change the local type of conductivity (in particular, to create lateral p-n junctions).