The relationship between so-called anomalous diffusion and end-of-range defects was considered. It was noted that a knowledge of the type and origin of those defects would permit an understanding of how they affected dopant diffusion. It was possible to discover under which experimental conditions their density could be minimized, and to develop physical models for dopant diffusion which took account of the behavior of end-of-range defects during annealing. It was pointed out that dislocation loops were efficient trapping sites for B, and could not be neglected when performing realistic simulations. During annealing, these defects increased in size and reduced in density via the emission and capture of Si interstitials. Thus, end-of-range defects could be seen as being reservoirs which were able to maintain a high supersaturation of free self-interstitials during dissolution. This point-defect supersaturation led to a marked increase in B diffusivity via the formation of excess Si(I)-B pairs. It was concluded that the experimentally observed macroscopic motion of a dopant was the result of 2 competing phenomena.

A.Claverie, L.Laanab, C.Bonafos, C.Bergaud, A.Martinez, D.Mathiot: Nuclear Instruments and Methods in Physics Research B, 1995, 96[1-2], 202-9