Physical modelling and optimization of ferroelectric FETs for memory applications.
The aggressive downscaling of nowadays electronics is not limited only to logic devices, but also to memory. In order to keep the pace preordained by Moore’s law the industry is currently in pursuit of the next generation devices to enable further scaling. The recent discovery of ferroelectric phases in doped high-k materials such as hafnium or zirconium oxide caused a major change in a memory device landscape and rekindled the interest in ferroelectrics. Ferroelectric Field-Effect Transistors (FeFET) are potentially faster and more energy efficient alternative to 3-D NAND, which place them as a viable candidate for storage class memory (SCM) devices. Nevertheless, the behavior and physical laws governing these materials are still far from being understood. Nowadays reliability issues such as fatigue, imprint and retention seem to be a crucial problem and providing a model of these phenomena would be benefitable for both industry and scientific community. The aim of this thesis is to develop a physics-based compact model of FeFET in order to enhance the understanding of the device, test it in representative benchmark circuits to study the viability of FeFETs in next generation memory circuits and investigate the electrical impact of degradation mechanisms.