Investigation of Extrinsic Grain Boundary Dislocations in Copper

The extrinsic grain boundary dislocations (EGBD's) in copper have been studied by transmission electron microscopy. It was shown that some EGBD's remain on grain boundaries after recrystallization and annealing at 0.6 T,, but not on those annealed at 0.8 T,. For several grain boundaries the crystallography of the defects was examined in detail and the dynamic theory of diffraction was used for computations of contrast profiles from EGBD's. The results were compared with observed images to obtain some information about the Burgers vectors of the analysed defects. In nearly all observed caGClines of EGBDFs lie along the intersections of ( I 1 1 ) planes with the boundkries plane, but the directions of Burgers vectors of these dislocations very often do not belong to the same { 111 } plane. Evidence was also obtained that EGBD's are able to react each other by a proces involving climb. These observations are regarded as evidence of the recovery of the grain boundary structure. 1 . Introduction. In the last ten years, one can again observe a phase of considerable growth of interest in structure of grain boundaries in metals. This time it is mainly stimulated by advance in the technique of transmission electron microscopy (TEM) and has provided information about various kinds of linear features in grain boundaries. 1t should be emphasized that TEM does not reveal the grain boundary structure directly by obtaining the images of individual atoms or lattice planes, or by producing the diffraction patterns from grain boundary structure (perhaps with the exception of a very few special cases). However, some contrast effects observed in the images of grain boundaries are interpreted as defects which by very careful analysis can provide some information about grain boundary structure itself. In our laboratory the research programme on grain boundary structure has been taken up recently with the emphasis on two. problems : a) analysis of the stability of extrinsic grain boundary dislocations during annealing fol l~wing primary ,recrystallization ; b) determination of Burgers vectors of these dislocations by comparing the computed intensity profiles with measured ones. The first problem is related to a postulated model of grain boundary recovery [I]. Accordingly to the model, during recrystallization lattice dislocations are transformed into grain boundary defects, the annihilation of which leads to the recovery of perfect grain boundary structure and takes place at temperatures above those heccesary for p i m i r y recrystallization. The model has been .proposed on the basis of measurements of grain boundary energy and it is thought that more conclusive information about the-process can be obtained by using electron microscopy. In this paper some results are presented concerning both problems. 2. Classification of grain boundary linear defects. As rather considerable differences' exist in terminology concerning various types of linear defects (or features) observed on grain boundaries, a . Glassification has been proposed [2-41 for the purpose of this work, Table I, to avoid the misinterpretation of experimental results. This classification 'is based on three criteria : 2 . 1 Character of elastic displacement field of defect. This leads to a distinction between grain boundary steps (GBS) and grain boundary dislocaArticle published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1975405 R. A. VARIN, J. A. KOZUBOWSKI AND M. W. GRABSKI