Silicon and Its Compounds with Defects and Dopants

As the term "Silicon Valley" reminds us, silicon in many forms is the major substrate upon which the empire of electronics has been built. Professors Banerjee and Hwang and their students are investigating the electronic properties of this material and its compounds in great detail, using computer analyses that rely upon density-functional theory (DFT), which enables a very exact representation of interactions at the atomic level. Their investigations will enable electrical engineers to design new and improved equipment to detect, amplify, and transform input signals, often over submicroscopic distances at extremely rapid rates, in what are known as nanosystems.

Banerjee and Hwang study "defects" that originate within silicon surfaces and bulk compounds (interstitial substitutions or vacancies in a crystalline silicon substructure). They also study "dopants"--atoms deliberately added to the substrate to change its electronic properties. Defects and dopants form, diffuse, and may be annihilated at silicon surfaces, at the interfaces between silicon and various oxides, and at the interfaces between amorphous and crystalline silicon. Defect-dopant complexes also diffuse, agglomerate, and precipitate within these systems.

Precise control of dopant concentration profiles during the fabrication of devices is dependent, Banerjee and Hwang are finding, on knowing more about formation, diffusion, clustering, annihilation, and precipitation of defects and dopants and their complexes within a wide variety of silicon-based systems. Their studies thus go beyond the usual theoretical studies of the fundamental behavior of defects or dopants in bulk silicon or on pure surfaces. "Nanoelectronics requires control of junction formation, often at ultrashallow depths," the investigators note, "and very little is known about the behavior of defect-dopant complexes at these depths, particularly within complex surfaces and interfaces or in amorphous regions. Hence the broad reach of our modeling studies."