Unusual twinning in annealed copper

Abstract

A N unusual thermal etch figure in copper is described and an explanation in terms of twinning is suggested. A flat polycrystalline specimen of OFHC copper cut from a rolled sheet was electro-polished in orthophosphoric acid and annealed at 950°C in a vacuum of approximately 10-• mm Hg for 2 hr. The specimen was found to be thermally etched, the markings on the surface being consistent with the assumption that they were the traces of the {111} planes which are developed preferentially because they have the lowest free energy. However, on one grain (the grain shown in Fig. 1) the traces formed an almost regular pentagon and at first sight it seemed unlikely that they could all be {111} traces. These traces were studied further because it was thought that they might be a type of growth spiral. A fringe pattern showed that this was not so, the grain being sensibly flat and the only features having steps were the striations and the faint curved line passing out from the center of the pentagons. To obtain some additional information, the specimen was deformed in tension. The effect of the stress was first to cause the appearance of lines running from the centers to the apexes of the pentagons and then to tilt the five segments so formed. At a much greater stress, appreciable slip appeared on the grain; these effects can be seen in Fig. 2. The micrograph was taken at a fairly late stage in the deformation and the tilting is quite marked. The magnification is the same as in Fig. 1, and although some slip can be seen it could be observed better by varying the illumination and using higher magnifications. In each segment of the grain, slip (which takes place on the { 111} planes) was parallel to the side of the pentagon in that segment. The marked changes in direction of slip from segment to segment seemed to be incompatible with the presence of a regular array of subgrains. These results suggest that some of the five segments were twins. If this is so, since the twinning plane in copper is the { 111} plane, the radial lines must be traces of the {111} planes on the surface. From the symmetry of the striations about the radial lines, these planes must be almost perpendicular to the surface. The angle between <111> directions in face-centered-cubic copper is 70°32'. If, in Fig. 3, OAB (triangle 1) is taken as matrix and OA, OB are { 111} planes viewed edge on, then AB is in the correct direction to be the trace of either of the other two { 111} planes on the surface. If triangle 2 is the twin (about the { 111} plane perpendicular to the surface and containing OB) of triangle 1, then once again it can be shown that all lines are traces of {111} planes. Similarly, triangle 3 is the twin of triangle 2, 4 of 3, and 5 of both 4 and 1. (The sequence is, of course, quite arbitrary.) However, if this were the case there would be a discrepancy of 7°20' to be accounted for. It is suggested that the faint curved line passing outward from the center of the pentagon represents either a R. L. SEGALL is associated with