No abstract

No abstract

Diffusion of platinum at low temperatures is a convenient way to characterize vacancy profiles in silicon. This article summarizes the experiments performed to find a standard procedure, discusses the pitfalls and limitations, and shows the applicability of the method. The results of experiments with FZ and CZ samples in the temperature range from 680 deg C to 842 deg C were found to disagree with the predictions of models published in the literature. Therefore, parameters governing the diffusion of point defects and platinum in silicon were determined for this temperature range

The formation of ultra-shallow junctions (USJs) for future integrated circuit technologies requires preamorphization and high dose boron doping to achieve high activation levels and abrupt profiles. To achieve the challenging targets set out in the semiconductor roadmap, it is crucial to reach a much better understanding of the basic physical processes taking place during USJ processing. In this paper we review current understanding of dopant-defect interactions during thermal processing of device structures – interactions which are at the heart of the dopant diffusion and activation anomalies seen in USJs. First, we recall the formation and thermal evolution of End of Range (EOR) defects upon annealing of preamorphized implants (PAI). It is shown that various types of extended defect can be formed: clusters, {113} defects and dislocation loops. During annealing, these defects exchange Si interstitial atoms and evolve following an Ostwald ripening mechanism. We review progress in developing models based on these concepts, which can accurately predict EOR defect evolution and interstitial transport between the defect layer and the surface. Based on this physically based defect modelling approach, combined with fully coupled multi-stream modelling of dopant diffusion, one can perform highly predictive simulations of boron diffusion and de/re-activation in Ge-PAI boron USJs. Agreement between simulations and experimental data is found over a wide range of experimental conditions, clearly indicating that the driving mechanism that degrades boron junction depth and activation is the dissolution of the interstitial defect band. Finally, we briefly outline some promising methods, such as co-implants and/or vacancy engineering, for further down-scaling of source-drain resistance and junction depth.