Diffusion in evaporated films of gold—lead

Abstract

Earlier measurements of reflectivity changes in evaporated two-layer gold—lead films by Schopper were explained in terms of a gradual precipitation of AuPb₂ which was equivalent to a gradual concentration change. Changes occurred only at the gold surface and these were attributed to diffusion of the lead atoms. New measurements have been made which show that the reflectivity changes at the gold surface are due to a moving phase boundary rather than a gradual concentration change, and that corresponding changes take place at the lead surface but no reflectivity changes occur simply because the reflectivity of the compound formed is the same as the reflectivity of pure lead. For the full reflectivity change to be produced the lead has to be at least 3•6 times as thick as the gold, and under these conditions the gold layer is completely converted to AuPb₂. These changes cannot be explained by the earlier theory which is strictly applicable only to perfectly miscible metals. Diffusion rates have been measured over a range of temperatures to obtain an activation energy of 17•1 kcal/mole ± 1•5% and the changes are attributed to diffusion of gold into lead with formation of the compound AuPb₂. This gives a diffusion rate D = 18•9 × 10⁻¹⁵ cm²/sec at 20°C and D ₀ = 0•11cm²/sec. No evidence of grain-boundary diffusion has been found and diffusion appears to be due to a vacancy mechanism.