On the annealing of quenched-in vacancies in gold

Abstract

Initially 99.999% pure gold was annealed in air at elevated temperatures and then quenched. The quenched-in vacancy resistivity annealing was then investigated over a wide range. Several specimens were strained after quenching to test the effects of straining on subsequent annealing. Correlated observations of accompanying vacancy precipitation were made using electron microscopy. It was found that: (1) the pre-quench annealing reduced the initial dissolved impurity content; (2) highly purified specimens exhibited annealing half times up to 10² longer than less purified specimens; (3) a large increase in the density of vacancy precipitates after annealing occurred with decreased purity; (4) the effective migration energy, E _eff ^m, was 0·71 ± 0·03 ev for vacancy concentrations between 1·5 × 10⁻⁶ and 5 × 10⁻⁵, for annealing temperatures between 30°c and 185°c, and for all purities; (5) post-quench straining increased the annealing rate but did not affect E _eff ^m. The results are explained by the nucleation of vacancy precipitates by impurities, thereby decreasing the number of jumps to annihilation without affecting E _eff ^m. The plastic straining added dislocation sinks and possible precipitation sites at dislocation-produced debris. Only limited conclusions can be reached regarding the identity of the migrating defects. It is suggested that the monovacancy migration energy is near 0·84 ev and that the observed 0·71 ev migration energy may correspond to a rather tightly bound divacancy with binding energy ε 0·3 ev.