• Cristiano, Fuccio

Abstract

This dissertation summarises my research activities in the field of Ion Implantation-Induced extended defects and of their impact on the properties of Ultra-Shallow source/drain junctions (USJs) in miniaturized MOS transistors. The most common method for the fabrication of source/drain regions consists in the localized doping of the substrate material by ion implantation, followed by thermal annealing to achieve electrical activation. The major problem related to the use of ion implantation is the formation of various defect types resulting from the precipitation of the large amounts of interstitials and vacancies generated during the implantation process and their interaction with dopant atoms during annealing. The various complex interactions between the defects and the implanted dopants are at the origin of the diffusion and activation anomalies that represent the major obstacles to the fabrication of USJs satisfying the ITRS requirements. The main results of my work will be presented in three parts. The first part is dedicated to the fundamental studies on the formation and evolution of implant-induced defects and on their impact on transient enhanced diffusion (TED). These studies contributed (i) to provide a unified description of implantation-induced defect evolution, explaining why, depending on the implant and annealing conditions, a given defect type is formed, dissolves during annealing or transforms into a larger defect with different crystallographic characteristics and (ii) to improve the existing models by extending them to all defect families, including a correct TED dependence on the defects' size distributions. In the second part, I will focus on the defect-dopant interactions causing dopant activation anomalies, due to their impact on the active dose and is some cases, also on the carrier mobility. In the case of p+-n junctions formed by Boron implantation, these anomalies are due to the formation of small Boron-Interstitial Clusters (BICs), which will be at the centre of all the studies presented in this part. Other investigated defect-dopant interactions include the formation of Fluorine-related Si interstitial traps, used to reduce both B Transient Enhanced Diffusion and dopant deactivation, and the dopant trapping by implantation-induced defects. The progressive introduction of advanced processes and materials in the semiconductor industry during the last decade raised some specific questions related to the fabrication of USJs, including the formation of implant-induced defects during ultra-fast annealing, their evolution in the presence of the buried Si-SIO2 interface in SOI materials or the Boron activation stability in Germanium. We will address these issues in the third part of this presentation. Due to the increased difficulties to maintain the MOS miniaturization pace (as well as to the approaching of its physical limits), the general context of the MOS-related research domain has largely evolved over the last years. On the one hand, the continuous optimisation of advanced doping and annealing schemes for the fabrication of USJs will therefore have to deal with the increasingly important requirement of reducing power consumption in future device generations. On the other hand, the years 2000s have seen the emergence of the so-called "More-than-More" domain, consisting in the addition of novel functionalities to electronic devices based on (or derived from) Silicon MOS technology. The perspectives of my research activity within this "extended-CMOS" context will finally be presented at the end of the presentation.

Topics

• impedance spectroscopy
• cluster
• mobility
• semiconductor
• Silicon
• precipitation
• Boron
• annealing
• activation
• interstitial
• Germanium