Techniques for Investigating Integrated Circuit Dielectric Isolation Media

Abstract

This paper describes two different techniques for investigating the dielectric isolation of integrated circuits manufactured by this method of isolation. The study concentrates on the two radiation effects of induced transient conductivity and charge storage in silicon dioxide. The first technique is theoretical involving a mathematical model which considers some of the physical processes in silicon dioxide under irradiation. A system of partial differential equations is derived from a consideration of some of the possible mechanisms and carrier kinetics occurring in the oxide as a result of radiation. A sub-routine was then written for an existing computer program (SCEPTRE) which was used to numerically solve the continuity equations describing the model. Primary outputs of the computer program are the transient current density, total charge in the oxide, and the surface charge induced in the silicon by the charge in the oxide which then allows one to also calculate the induced charge in the gate electrode. In addition, available as outputs are the local electric field, local charge density, and the instantaneous density of electron and hole charge carriers throughout the dielectric. The other technique is an empirical method of nondestructively determining an existing nominal electric field in unirradiated, irradiated and then annealed silicon-silicon dioxide-polycrystalline silicon (SOS) structures. This method of investigation is based on the quadratic electro optical Kerr effect. From results obtained by applying external bias voltage on these SOS-structures it was concluded that the nominal internal electric field in the oxide is directed toward the polycrystalline silicon region.