Abstract
From a device structure standpoint it would be advantageous to sandwich laterally defined features between layers of epitaxially grown material. In silicon this is commonly dope by growing the bottom layer, patterning the desired feature, and growing a second layer. Unfortunately, this process has not been practical in GaAs for the same reason that there is no true MOS technology in GaAs: The. GaAs surface is irreparably damaged when it is exposed to the atmosphere leading to the formation of undesirable interface states. Heterojunction FET's are feasible only because high quality epilayers are grown during a single run in an ultrahigh vacuum environment. Standard growth methods allow for variation of doping and material content only in one direction, normal to the wafer surface. Varying the material in more than one dimension without the use of prohibitively exotic equipment requires removal of the wafer from the growth apparatus for lateral processing between material growths. Thus the problem that this thesis attempts to address: How to protect a GaAs surface during a lateral processing step and initiate regrowth leaving behind an electrically invisible restart interface. The potential applications of the development of a successful interrupted growth scheme for GaAs are numerous and far reaching. Specifically it would allow the fabrication of advantageous device geometries that are not possible under single material growth runs. Although this thesis deals exclusively with ion implanted interrupted growth by Molecular Beam Epitaxy, some of the concepts arid theories can be extended to other growth methods. It is both a review of previous work and a report of our attempts at Purdue to fabricate the first interrupted growth HIGFET's and MISFET's. Mechanisms behind the success and failure of GaAs interrupted growth are discussed and several experiments involving passivation materials and new interrupted growth schemes are proposed
Date of this Version
1-1-1988
Comments
This Work Supported By: Indiana Corporation for Science and Technology contract: P-5058