I. Introduction
To improve the performances of electronic devices, the size of their active components is scaled down according to the International Technology Roadmap for Semiconductors (ITRS) [1]. In this context, the current bulk complementary metal–oxide–semiconductor FETs are evolving toward nanoscale ultrathin body silicon-on-insulator (SOI) structures, which suffer less from short-channel effects and offer steeper subthreshold slopes [2]. This favorable behavior is attributed to better electrostatic control obtained by reducing the thickness of the silicon body on top of the buried oxide [3]. At the same time, the thickness of the traditional gate dielectric material has been scaled down below 2 nm [4]. At this size, the gate leakage currents can reach a value of 10 [5] due to their exponential dependence on the oxide thickness. They may become the dominant leakage mechanism, deteriorate the device reliability, and cause most of the power consumption. To circumvent this problem, is replaced by “high-” dielectrics as or [6] with an equivalent oxide thickness (EOT). These materials reduce the tunneling leakage due to the larger physical thickness but provide the same gate capacitance due to their higher permittivity. However, they also exhibit an increased interfacial trap concentration as compared to and cause a degradation of the channel mobility [7]. Hence, the high- materials are often used in combination with a thin layer, forming a so-called gate stack.