Bonding constraint-induced defect formation at Si-dielectric interfaces and internal interfaces in dual-layer gate dielectrics

Abstract
As aggressive scaling of integrated circuits continues into the next century, insulators with dielectric constants higher than SiO2 with different local bonding arrangements will be required to increase gate dielectric capacitance in field effect transistor devices. An important issue in semiconductor device physics is determining whether differences between the bonding at (i) Si–SiO2 interfaces and (ii) interfaces between crystalline Si and alternative gate dielectric materials will result in increased densities of electrically active defects at the alternative dielectric interfaces, thereby limiting targeted levels of performance and reliability. In particular, it is important to understand from a chemical bonding perspective why Si–SiO2 interfaces display both low defect densities and high reliability, while other interfaces such as Si–Si3N4 with similar bonding chemistry, display defect densities that are at least two orders of magnitude higher. Building on previously established criteria for formation of low defect density glasses and thin films, constraint theory is extended to crystalline Si-dielectric interfaces that go beyond Si–SiO2 through development of a model that is based on the average bonding coordination at these interfaces. This approach identifies quantitative bonding criteria that distinguish between device-quality and highly defective interfaces. This extension of constraint theory is validated by its application to interfaces between Si and stacked silicon oxide/nitride dielectrics which demonstrates that as in bulk glasses and thin films an average coordination, Nav>3 yields increasingly defective interfaces. Finally, the universality of this application of constraint theory is demonstrated by showing that defect densities scale with overcoordination in the same way in thin films and at interfaces.

This publication has 26 references indexed in Scilit: