816 



THE BELL SYSTEM TECHNICAL JOURNAL, OCTOBER 1951 



Figure 15 shows electron diffraction patterns obtained from the several 

 replica films, indicating that crystallite size ranges from greater than 100 A 

 for aluminum down to almost amorphous for germanium and amorphous 

 for silica. 



/■ 



ALUMINUM 



GOLD-MANGAHIN 



CHROMIUM 



GE, AFTER HEATING 



GE, AS EVAPORATED 



SILICA 



Fig. 15 — Electron dilTraction patterns of evaporated materials, in order of decreasing 

 crystallite size. The aluminum pattern is due to crystallites of about lOOA. The gold- 

 manganin pattern shows at least two sharp rings not due to gold and diffuse bands due 

 to very small crystallites with the structure of gold. The chromium pattern is due to 

 very small crystallites of Cr. The pattern of a germanium replica film after heating in 

 air to 390° shows partial recrystallization, with rings due to quite small crystallites super- 

 posed on the amorphous pattern of the Ge as evaporated. 



4. Interpretation of Micrographs of Replicas 



The electron image pattern due to the repHca is reproduced by the ex- 

 posure of a photographic plate. The complex problems associated with 

 photographic reproduction^^ cannot be discussed here. For a number of 

 reasons, blackening of the plate is not linearly related to repHca film thick- 

 ness. The number of electrons scattered out of the beam* is proportional 

 to thickness only as a first approximation, the blackening vs. exposure 

 curve of the plate is not linear, and there is also a roughly uniform back- 

 ground due to inelastically scattered electrons. At the lower electronic 

 magnifications, field distortion is also a factor affecting local intensity in 

 the electron image. Furthermore, the geometrical relation between the 

 electron beam incident upon the thin mesh-supported film and the atom- 

 beam is usually not known accurately, and indeed may vary locally over 

 the replica in the manner of which Fig. 9 is an extreme example. In con- 



" W. T. Wintringham, Proc. I. R. E., 38, 1284 (1950). 

 ♦Ref. l,ch. 19orref. 7, p. 541. 



