Direct Observation of Dislocations and Their 

 Movement in Metal Foils 



P. B. HiRSCH, R. W. HoRNE, and M. J. Whelan 



Cavendish Laboratory, Cambridge 



In order to explain the low values of the shear 

 stress required to start plastic flow in metal crystals, 

 it is necessary to postulate the existence of a lattice 

 imperfection, known as a dislocation. There is a 

 considerable amount of indirect evidence from etch- 

 ing and precipitation experiments (9, 13) that such 

 imperfections exist in metal crystal, whilst for inor- 

 ganic crystals such as AgBr and NaCl (1,5) the 

 dislocation networks may be decorated by various 

 techniques and made visible under the optical micro- 

 scope. However no similar direct observations have 

 been made on metals, simply because it is not possible 

 to examine interior structures with optical tech- 

 niques. It occurred to the authors that much useful 

 information on the arrangement and movement of 

 dislocations in metals might possibly be obtained 

 by examining thin foils directly in the electron 

 microscope. Electron optical transmission experi- 

 ments with gold foils had shown that the contrast 

 was essentially due to Bragg scattering which is 

 structure sensitive. It was therefore thought that dis- 

 locations might be made visible by virtue of their 

 strain fields. This paper is a short account of some 

 observations on dislocation distributions and move- 

 ment in aluminium foils. A fuller account is published 

 elsewhere (8). 



Aluminium was used in this work on account of its 

 transparency to electrons, and also because much is 

 already known about the substructures formed by de- 

 formation in this metal. Beaten foils \ /i thick, either cold 

 worked or annealed at 350 C in vacuo, were etched in 



dilute hydrofluoric acid. The foils were examined directly 

 in the Siemens and Halske "Elmiskop 1" operating at 

 80 KV at an instrumental magnification of 40,000 -< . 



Evidence for the visibility of dish^cation lines. — 

 Figs. 1-5 show some typical micrographs obtained. 

 The following facts leave little doubt that individual 

 dislocation lines are being observed. 



(a) Fig. 1 shows that the specimens contain a 

 substructure of subgrain diameter about 1 /<. The 

 misorientations across the boundaries have been 

 determined by diffraction experiments and are found 

 to be about I' or 2°. Fig. 2 shows a typical boundary 

 under higher magnification, showing that it is pos- 

 sible to resolve individual dislocations in the boun- 

 daries. An interpretation in this case is that this is a 

 simple (110) tilt boundary consisting of edge disloca- 

 tions traversing the foil. MicrodiflFraction patterns 

 show that the foil has a strong preferred orientation 

 with a [100] cube axis normal to its surface. This 

 orientation is favourable for the formation of (110) 

 tilt boundaries normal to the foil. Examples are 

 observed of subgrain boundaries traversing the foil 

 at an angle, in which the three dimensional character 

 of the dislocation networks is evident. 



(b) The dislocation density estimated from several 

 micrographs similar to fig. 2 is about 10'" per sq.cm. 

 This is in good agreement with previous estimates 

 from x-ray evidence (4, 7), and with the average 

 misorientation angles 1 or 2" observed from diffrac- 

 tion patterns. 



Fig. 1. Siibgrains in cold beaten Al. The average subgrain 

 size is about 1 //, the average angular misorientation about 

 1^°. The dislocation density is lO^" per sq.cm. Extinction 

 contours due to large range strains are shown at A and B. 

 Magnification 12,000. 



Fig. 2. A sub-boundary consisting of uniformly spaced dis- 

 locations. The average spacing of the dislocations is about 

 175 A. Magnification 120,000. 



