It was recalled that molecular dynamics studies had shown that significant clustering of vacancies and self-interstitial atoms took place by the end of the quenching stage of cascades, and that small interstitial clusters were glissile; with migration energies of the order of 0.1eV. The spatial segregation and clustering of vacancies and self-interstitial atoms led to the differential production of mobile vacancies and self-interstitial atoms. This exhibited a marked temperature dependence. At temperatures above recovery stage V, vacancies could evaporate from clusters. Meanwhile, large small interstitial clusters which were produced in the cascade remained stable; thus leading to a differential increase in the numbers of mobile vacancies. It was suggested that such a so-called production-bias might be responsible for void swelling. Simulated annealing was used to model the differential production of mobile vacancies and self-interstitial defects, as a function of temperature, for isolated 25keV cascades. These results indicated that, above stage-V (where the cascade vacancy clusters were unstable), some 80% of the post-quench vacancies escaped from the cascade volume while some 50% of the post-quench self-interstitials remained in clusters. The results were sensitively dependent upon the relative fractions of self-interstitials which remained in small mobile clusters and large sessile ones. These fractions were suggested to depend upon the cascade energy.

H.L.Heinisch, B.N.Singh: Journal of Nuclear Materials, 1996, 232[2-3], 206-13