Fundamental Defect Complexes and Nanostructuring of Silicon by Ion Beams

Abstract

Silicon technology has become a cornerstone for the technological advances in our society for the last five decades. For carrying on with minimization of electronic devices a huge effort has been directed toward the technological development and fundamental understanding of physical processes associated with the ion implantation into silicon. Indeed, in ion implantation the impurities are intentionally introduced into a matrix lattice with the help of accelerated ion beams selectively modifying the properties of the implanted area. In addition to the introduction of the doping impurity the penetrating ions create defects that can be electrically active potentially affecting the device performance. In spite of a long research activity in the field there are still several open fundamental questions remaining, and this thesis contributes to the understanding of ion implantation induced defect complexes in silicon. Firstly, we have studied the electrical properties of vacancy type point defect complexes generated in single collision cascades during heavy ion bombardment of silicon. Because of a high generation rate of defects within the “ion track” regions, a characteristic pattern of nanochannels having modified Fermi levels due to the local compensation around each ion trajectory is formed in n-type Si. The phenomenon has been studied using spectroscopic and imaging techniques, specifically deep level transient spectroscopy (DLTS) and scanning capacitance microscopy (SCM). The SCM measurements show a characteristic random pattern of reduced SCM signal correlated with the density of the ion impacts. Moreover, a strong correlation is detected between the probing frequency and the emission rate of the single negative acceptor level of the divacancy V2(-/0) in Si. Further, DLTS reveals a significant filling time increase for all electronic levels originated from vacancy complexes with increasing ion mass as probed within the ion track regions. The results of isochronal annealing studies of vacancy complexes generated by heavy ion implants are also explained in terms of the revisited local compensation model. An improvement of the model is proposed, where the divacancy is considered to be available in two fractions; (1) highly localized centers along the core track regions V2dense and (2) centers located outside ion tracks V2dilute. The relative abundance of V2dense/V2dilute is ion mass dependent. In this model the V2dense fraction does not contribute to the doubly negative divacancy V2(=/-) signal due to local carrier compensation, and the DLTS amplitude of V2(=/-) is determined only by the V2dilute fraction. Our finding clarifies a long lasting discussion in literature on the DLTS amplitude difference between V2(-/0) and V2(=/-) in ion implanted n-type Si. Secondly, the thesis contains an investigation of the dominant electron trap in p-type Si (Ec -0.25eV), where Ec is the conduction band edge. The Ec - 0.25eV trap has previously been ascribed to the boron interstitial-oxygen interstitial (BiOi) complex, but our study shows no oxygen and only a weak boron dependence on the intensity of the level, challenging the BiOi identification. Finally, the thesis explores the use of defect engineering by introducing nanosized vacancy clusters (cavities) when synthesizing buried SiO2 by ion implantation (SIMOX). Scanning spreading resistance microscopy measurements show that oxide nucleation can be enhanced by introducing cavities, potentially reducing the required oxygen dose during the SIMOX processing.