Wafers were implanted with Si+, to doses of 2 x 1014 or 1015/cm2, and were annealed (700 to 1000C, 0.25 to 16h) with or without a SiO2 cap. Plan-view transmission electron micrography revealed the densities of interstitials which were trapped in (311) defects and type-I and type-II perfect loops. The results showed that, after the lower dose (below the amorphization threshold), the predominant defect at 700C was the (311) defect; with a much lower concentration of type-I loops. The total trapped interstitial concentrations in both types of defect was of the order of 7 x 1013/cm2 after annealing (700C, 1h). The (311) defects began to dissolve after several hours at 700C, but their dissolution rate was lower than that previously reported. The reduced dissolution rate was not changed by capping. The type-I loops exhibited some growth during (311) dissolution, but less than half of the released interstitials appeared to be trapped by the type-I loops. In the case of the higher-dose samples, amorphization occurred and both type-II (end-of-range) loops and (311) defects were observed after annealing at 700C. The total number of trapped interstitials after annealing at 700C for 1h was also about 7 x 1013/cm2. However, the ratio of (311) defects to loops had interchanged such that the predominant defect was the type-II loop. During annealing, the (311) defects again exhibited a reduced dissolution rate, and the type-II loops were in the growth regime. An increase in the annealing temperature to 800C resulted in further growth of the type-II loops, and all of the (311) defects either dissolved or unfaulted. The growth of the type-II loops appeared to be greater than could be accounted for by the (311) defects. There was also a high level of strain in the lattice that could not be accounted for by the (311) defects. These results implied that there might be an additional source or interstitials, besides (311) defects, which was involved in amorphization.

K.S.Jones, J.Liu, L.Zhang, V.Krishnamoorthy, R.T.DeHoff: Nuclear Instruments and Methods in Physics Research B, 1995, 106[1-4], 227-32