Transient Enhanced Diffusion In Silicon Research Articles

Modeling of the effect of the buried Si–SiO2 interface on transient enhanced boron diffusion in silicon on insulator

The effect of the buried Si–SiO2 interface on the transient enhanced diffusion (TED) of boron in silicon on insulator (SOI) structures has been investigated. To this purpose, boron marker layers were grown by chemical vapor deposition on Si and SOI substrates and implanted under nonamorphizing conditions with 40 keV Si+ ions. The experimental results clearly confirm that the Si–SiO2 interface is an efficient trap for the Si interstitial atoms diffusing out of the defect region. Based on these experiments, existing models for the simulation of B TED in silicon have been modified to include an additional buried recombination site for silicon interstitials. The simulation results provide an upper limit of ∼5 nm for the recombination length of interstitials at the Si–SiO2 interface.

Implant temperature dependence of transient-enhanced diffusion in silicon (1 0 0) implanted with low-energy arsenic ions

The diffusion of arsenic implanted into silicon at low ion energies (2.5 keV) has been studied with medium-energy ion scattering, secondary ion mass spectrometry and four-point probe measurements. The dopant redistribution together with the corresponding damage recovery and electrical activation produced by high-temperature (550–975°C) rapid thermal anneals has been investigated for a range of substrate temperatures (+25, +300 and −120°C) during implant. Initial results show an implant temperature dependence of the damage structure and arsenic lattice position prior to anneal. Solid-phase epitaxial regrowth was observed following 550°C, 10 s anneals for all implant temperatures and resulted in approximately 60% of the implanted arsenic moving to substitutional positions. Annealing at 875°C resulted in similar arsenic redistribution for all implant temperatures. Following annealing at 925°C, transient-enhanced diffusion was observed in all samples with more rapid diffusion in the +25°C samples than either the −120 or +300°C implants, which had similar dopant profiles. In the 975°C anneal range, similar rates of implant redistribution were observed for the +300 and +25°C implants, while diffusion in the −120°C sample was reduced. These observations are discussed qualitatively in terms of the nature and density of the complex defects existing in the as-implanted samples.