Samples of n-type and p-type float-zone material, which contained 1018/cm3 of Sn, were irradiated with 2MeV electrons to a dose of 1018/cm2 and were examined by means of electron paramagnetic resonance. The p-type material yielded only the well-known Si-Ge-G29 signal, due to the Sn-vacancy complex. The as-irradiated n-type material also exhibited the Si-G7 signal (DK4); which had been attributed to V2- in a set of slightly inequivalent configurations. There was also a signal, DK1, which arose from defects with S = 1/2 that contained one Sn nucleus. The DK1 signal, which was attributed to (SnV-V)-, underwent a reversible triclinic-monoclinic transformation at about 15K. Annealing at 428K removed the Si-G29 and DK4 signals and produced a 6-fold increase in Si-G7 and DK1. The kinetics indicated that about 50% of SnV was transformed into V2 and (SnV-V). Annealing at 503K destroyed Si-G7 and DK1; with the decay of the latter being linked to the appearance of 2 new signals: DK2 and DK3. These arose from defects with S = 1/2, monoclinic-I symmetry, and 2 equivalent Sn nuclei each. They were identified as being (SnV-VSn)- and (Sn2V-V)-. The structures of the Sn-divacancy complexes were explained in terms of the modifications which were imposed on the basic divacancy structure by the larger size and lower ionization potential of the Sn atoms as compared to Si atoms. A model was proposed, for the migration of (SnV-V) at 500K, which involved the process: (SnV-V) + Sn → (SnV-VSn).

Divacancy-Tin Complexes in Electron-Irradiated Silicon Studied by EPR M.Fanciulli, J.R.Byberg: Physical Review B, 2000, 61[4], 2657-71