|Thesis abstract: |
Flash memory is today the leading solid-state non-volatile memory technology, allowing high integration density, low costs and good reliability. The continuous scaling has been the main driver of the success of this technology, pushing it, however, to its physical limits: the reduction of the array pitch is today limited by the increasing capacitive coupling among adjacent cells, the low number of electrons controlling cell state is raising issues related to the discrete nature of the charge flux from/to the floating gate, single electrons stored in the tunnel oxide result into more and more severe threshold voltage instabilities during read and data retention. Aim of this thesis is to study the emerging physical mechanisms limiting the reliability of ultra-scaled Flash memories, highlighting from a theoretical standpoint the fundamental limitations to the functionality of nanoscale memory arrays. All the work has been carried out with a scaling perspective, trying to assess the ultimate scaling limitations and to propose feasible solutions able to extend the success of the Flash technology to the future technology nodes.