Variation of end of range defects density with ion beam energy and dose: Experiments and simulations

Abstract
When the amorphous layer created by ion implantation becomes very thin, the density of End Of Range defects (EOR), formed after annealing, decreases. Some investigations [1] have attributed this fact to the motion, by glide or climb, of EOR defects towards the surface. To shed light on this point, many Ge⁺ implantations were carried out at different energies and doses, and very thin layers were elaborated. Based on XTEM observations correlated to "Excess Interstitials model", we show clearly that the observed phenomenon, can be rigorously explained through collisional arguments in the whole bottom of the crystal. When the beam energy or dose are lowered, the amorphous layer becomes thinner. So, the number of excess interstitials left beneath the c/a interface, which will cluster to form EOR defects, decreases. On the other hand, when EOR defects are formed at the surface's proximity (depth less than 75nm), they act as a sink of Si atoms and lead to a more pronounced reduction of super saturation, which will form EOR defects. However, no EOR defects are observed when the number of excess interstitials becomes less than a minimum threshold, which depends on the depth of these defects.