It is well known that in internally strained metals, for example, in cold, worked metals, the potential barrier which an atom must surmount to become mobile is very much lowered. The mutual diffusion of metals, therefore, must also become very rapid in such a strained metal. Dr. Matano was perhaps the first who had investigated this phenomenon quantitatively and systematically. His investigation, however, contains two defects : (1) he did not define the degree of plastic deformation, and (2) he used an alloy (Cu-Sn about 5 atomic%) which was in the concentration range where diffusion constants could not be considered as independent of concentration at low temperatures, say below 550°. This is due to the fact that the activation energy of this alloy decreases markedly towards high concentrations. The present author, taking care of these points, has carried out strict experiments. The alloy used is Cu-Al containing 15•7 atomic % of Al. On its surface pure. copper is electro-plated. In-the case of this alloy diffusion constants can be considered to be roughly independent of concentra-tion, especially at low temperatures, and the low temperature range is of importance, as we shall see later, for this investigation. Two kinds of base alloys are used : the one completely annealed, and the other cold-worked (elongated about 50 %). The method of experiments is almost the same as that of Dr. Matano, , There diffusion constants are decided from lattice constants which are proportional to alloy concentrations. The measurement of lattice constants, however, have been carried out more strictly using pure metallic silver as a standard sub-stance. The lattice constant of standard silver was, of cource, strictly measured previously. The results are tabulated in Table 2, and are illustrated in Fig, 2.
We can see from these results that the effect of internal strain disappears in the temperature range higher than a certain definite temperature (in this case about 770°). This is accounted for by the abrupt recrystallization, which changes the strained sample into annealed one instantaneously. The activation energy in the case of annealed alloy is 42, 900 cal/mol, which coincide completely with the value of Rhines and Mehl. And the activa-tion energy in the case of cold-worked alloy is 24, 600 cal/mol. These results will furnish some key points to elucidate the phenomenon of so-called “Platzwechselvorgang”.