It was pointed out that a reduction in the end-of-range defect density, seen after annealing relatively thin amorphous layers that had been created by ion implantation, could not be fully explained by the motion of dislocation loops towards the surface. The thickness of the amorphous layer was instead attributed to a reduction in the ion beam energy. The observations could be easily explained by collisional arguments. That is, when the beam energy was decreased, there was a decrease in the number of self-interstitials which remained beneath the crystalline/amorphous interface, survived complete recombination with vacancies, and clustered to form end-of-range defects. It was found that there was a minimum threshold, beneath which the supersaturation of interstitials was too low to promote clustering on (111) planes.

L.Laanab, C.Bergaud, C.Bonafos, A.Martinez, A.Claverie: Nuclear Instruments and Methods in Physics Research B, 1995, 96[1-2], 236-40