Introduction
Motivation for implementing high dielectric constant gate dielectric for advanced CMOS scalingSemiconductor devices need to have good performance, with a low cost and low power dissipation. For decades, research and development of semiconductor processing technology and device integration have focused on enhancing performance and reducing costs using SiO 2 as the gate dielectric and doped polysilicon as the gate electrode. The most effective way to enhance performance and reduce costs is to scale the device gate length and gate oxide. Scaling the gate length results in fabricating more devices per wafer (i.e., increase the device density) and thus reduce the cost per chip, while scaling the gate oxide enhances the drive current and reduces the short channel effects due to gate length scaling. However, as the gate oxide becomes thinner, the power to operate transistors increases because of greater gate oxide leakage current. To resolve this high gate oxide leakage problem, the mechanism of the carriers tunneling through the gate dielectric must be better understood. In an ideal metal-insulator-semiconductor (MIS) device, the current conduction in the insulator should be zero. In a real MIS device, however, current can flow through the insulating film by various conduction mechanisms. The two primary conduction mechanisms for electron tunneling through high quality gate dielectric are discussed below.
Direct tunnelingIn a metal-insulator-semiconductor (MIS) stack, when the oxide voltage (V ox ) is smaller than the metal-insulator barrier height, the electron tunnels directly from the metal electrode into other semiconductor electrode through the insulator. This is known as the direct tunneling process. The equation for direct tunneling current density (current normalized by device area) is proportional to