First principles computational studies of high-k dielectric stacks for next generation transistors

The usage of metal-SiO₂-Si gate stacks has been pervasive in microelectronic devices for the last four decades. Since continued device miniaturization through physical scaling has reached an impasse, the community is poised for the replacement of SiO₂ by HfO₂ based "high-K" dielectrics. However, widespread usage of the new high-K gate stack will be hampered due to several troubling issues, attributed to intrinsic O defects in HfO₂, as supported by experiments performed at varying levels of O exposure and anneals. In particular, positively charged O vacancies appear to be the most significant and mobile native defects contributing to such unexpected and undesirable behavior.