Sources of Thermally Generated Vacancies in Single-Crystal and Polycrystalline Gold

Abstract

The rate at which the equilibrium concentration of vacancies is generated during pulsing to elevated temperatures was studied in gold. Thin single-crystal slabs were gas-pulsed (\ensuremath{\sim}1250\ifmmode^\circ\else\textdegree\fi{}C ${\mathrm{sec}}^{\ensuremath{-}1}$) into the range 875-920\ifmmode^\circ\else\textdegree\fi{}C, held for short periods of time, and then down-quenched in order to trap the generated vacancies. The vacancies were detected by precipitating them as observable vacancy tetrahedra. Also, polycrystalline foils (grain radius \ensuremath{\sim}150-175 \ensuremath{\mu}) were electrically pulsed (\ensuremath{\sim}18\ifmmode\times\else\texttimes\fi{}${10}^{3}$ \ifmmode^\circ\else\textdegree\fi{}C ${\mathrm{sec}}^{\ensuremath{-}1}$) to 653 and 878\ifmmode^\circ\else\textdegree\fi{}C. In this case the generated vacancies were detected by their resistance at 4.2\ifmmode^\circ\else\textdegree\fi{}K. The densities of all possible vacancy sources (specimen surfaces, grain boundaries, subgrain walls, and free dislocations in the usual 3-dimensional network) were measured in the pulsed specimens. Knowing the shape of the thermal pulse, the diffusion problem of the maximum possible inward flux of vacancies from the various sources was solved and compared with experiment. A temperature-dependent monovacancy diffusion coefficient and time-dependent equilibrium boundary conditions at the sources were assumed. The results indicated that (1) free dislocations were the predominant sources; (2) the average free dislocation operated as an efficient source. In fact, the specimen filled up at a rate which was as large as 0.1 to 0.2 of the rate calculated on the assumption that vacancy equilibrium was maintained at all times along all free dislocation cores.