Analysis and modeling of total dose effects in advanced bulk CMOS technologies

The growing need to create accurate models for radiation effects (e.g., edge leakage and device-to-device leakage) that can be implemented in circuit simulators is necessary due to the increased use of commercial deep-submicron technologies in integrated circuits operating in harsh radiation environments. For example, most space electronics, medical-implantable devices, weapons systems, and radiation/accelerator facility instrumentation require a certain level of hardness to total ionizing dose (TID) and/or a complete understanding of the ionizing radiation response in the intended environment. When integrated circuits are exposed to ionizing radiation, the resulting effects can cause significant degradation to the performance of the circuit. Previous studies have demonstrated that in modern bulk CMOS technologies the most sensitive region to total dose damage is the shallow trench isolation (STI) oxide.;In this dissertation, experimental data and numerical simulations of field oxide field effect transistors (FOXFETs) and standard CMOS n-channel transistors are presented. The results are used to characterize the general susceptibility of STI oxides to radiation damage and to assess how degradation in the STI dielectric impacts CMOS device and circuit specifications. Through the use of closed form functions and reasonable estimates of critical technology variables (such as STI geometry and sidewall doping), the effects of ionizing radiation on bulk CMOS devices are also incorporated into analytical models. In the model, defect distributions in radiation sensitive STI oxides are analytically calculated as a function of time, integrated into radiation-aware surface potential equations, and then inserted into modified charge sheet equations to reproduce radiation-induced edge and device-to-device leakage currents. The results of the modeling approach are compared to experimental data obtained on 90 nm CMOS technology test structures, as well as to data obtained from radiation-enabled technology computer aided design (TCAD) simulations. This surface potential-based approach provides a methodology for analytically modeling the total dose response of modern MOS devices and circuits as a function of technological parameters, and suggests a path for the insertion of radiation-enabled compact models into circuit simulators.