A new scenario for ‹c› vacancy loop formation in zirconium based on atomic-scale modeling

Two-dimensional vacancy platelets as precursors for basal dislocation loops in hexagonal zirconium

“…The fraction of <c> loops is too small to argue that the remaining vacancy <a> loops all coalesced to form <c> loops 29 , or that the interaction between the irradiation cascades and a high density of pre-existing <a> vacancy loops accelerated <c> loop formation 5 . While MD simulations have shown that the small size <c> loop can transform from a 3D pyramidal defect, of which the critical size consists of 400 vacancies 30 , no such 3D pyramidal defects are observed in our experiment and other studies. Here, we observe a high density of 2D TVPs (one to three atomic layers thick) on the basal plane with sizes no more than 11 nm.…”

“…They can make up to 50% of the <a> loops at temperatures between 400 °C to 450 °C 3 and up to 70% at irradiation temperatures >450 °C 1 . From atomic-scale modeling, a mechanism for <c> loop generation based on the collapse of irradiation-produced pyramids has been proposed 30 . Unlike the prior proposals, the formation of <c> loops in this case would be independent of the formation of <a> vacancy loops, and would instead require the collapse of a stacking fault pyramid that has grown beyond a critical size 30 – 32 .…”

“…From atomic-scale modeling, a mechanism for <c> loop generation based on the collapse of irradiation-produced pyramids has been proposed 30 . Unlike the prior proposals, the formation of <c> loops in this case would be independent of the formation of <a> vacancy loops, and would instead require the collapse of a stacking fault pyramid that has grown beyond a critical size 30 – 32 . However, to date, these irradiation-induced pyramidal defects in Zr have not been observed.…”

“…The absence of vacancy <c> loops less than 10 nm on the basal planes and the presence of only interstitial <a> loops on the prismatic plane cannot explain the observed contraction along the c-axis of the irradiation growth phenomenon 1,11 . Being no smaller than 10 nm has suggested that the <c> loops require an incubation period in which point defects must cluster to a critical size before a <c> loop can be produced. The possible mechanisms for <c> loop formation have received much attention, and over the years, several hypotheses have been proposed 5,[29][30][31][32] . The formation of the <c> loops has been associated with the coalescence and subsequent collapse of <a> vacancy dislocation loops 5 .…”

Systematic development of ab initio tight-binding models for hexagonal metals

“…Atomistic simulations of most hexagonal metals and their alloys have been dominated by two classes of models: density functional theory (DFT) [1] and classical empirical potentials [most notably embedded atom model (EAM) [2]]. DFT models are effectively parameter free and transferable for use over a large range of different systems [3]; however, they are severely limited by the computational resources required for a simulation, with system sizes only rarely exceeding 1000 atoms [4][5][6]. Empirical potentials, on the other hand, require extensive parametrization, which may need to be redone with the addition of new elements to the model, and are generally only applicable for a limited number of problems.…”