INORGANIC REPLICATION IN ELECTRON MICROSCOPY 



805 



da = 60°. The plastic film was then ''floated off" on water and picked up 

 on 200-mesh screen; the plastic and particles were successively dissolved, 

 leaving the siHca replica.f The micrographs clearly show (a) shadow-edges 

 and (b) a fihn in the shadows. The enlargement shows a shadow within 

 which there exists a hole in the film and replication of completely shadowed 

 particles by the diffusing component. It follows from (a) that a part of the 

 incident material must stick where it strikes, and from (b) that a part 

 must diffuse into the shadows, somewhat in the manner diagrammatically 

 illustrated in Fig. 4. Densitometer traces through shadow edges show that 

 the film thins down a little as the edge is approached from the unshadowed 

 region, drops more or less abruptly at the edge, and continues to thin down 

 within the shadow^ as illustrated. Now the diffusing part must finally con- 



diffusing 



SHADOW EDGE ORIGINAL SURFACE 



Fig. 4 — Diagram illustrating diffusion of silica into shadowed regions. 



dense, and it is natural to assume that at each collision with the surface 

 the probabihty of sticking is a, and of diffusing is (1 — a), and that the 

 average molecule travels between collisions a small distance. Analysis of 

 densitometer curves on these assumptions leads to a value of a of about J, 

 and a range of about ^ /x, with a rather wide spread of values. ^^ However, 

 these assumptions would require the film to be vanishingly thin at distances 

 more than 3 ju from the shadow-edge. In fact, an extremely thin film is 

 found at even greater distances. The probability of sticking upon collision, 

 and the distance a molecule moves as a two-dimensional gas molecule 

 between colhsions with the surface, thus must be assumed to depend on 

 such factors as angle of impingement, energy of molecule, and perhaps 

 the nature of the surface. This latter initially is a plastic, probably covered 



t In particle study, a second evaporation of silica 180° in azimuth from the first pro- 

 duces a "thin-shell" replica more suitable for stereoscopic study of the particles. 

 18 C. J. Calbick, //. App. P/tys., 19, 119 (1948). 



