ELECTKON IMICKOSCOI'Y 



BtiM moil 

 iixomoii (na 



OBJECTIVE AND 

 PBOJECTOB i£Ha;s 



Fig. 1. Schematic diagram of a reflection elec- 

 tron microscope. 



projector lenses to form an image of the 

 surface. 



Factors governing the choice of di and 



$2 . The first reflection electron microscope 

 built by Ruska in 1933 had di -\- 62 = 90°. 

 The image was faint and the resolution only 

 about 5000 A. Von Borries (1940) had better 

 results using glancing angles of illumination 

 and viewing {di and 62 about 4° each). The 

 potentialities of the technique were not at 

 that time fully explored, but after the war 

 a number of workers developed the method 

 further. These included Kushnir, Biberman, 

 and Levkin (1951), Menter (1952), Fert 

 and Saporte (1952), Haine and Hirst (1953), 

 Cosslett and Jones (1955). In all this work 

 the von Borries arrangement of glancing 

 angles of incidence and viewing was used — 

 values of Oi of about 1° and 62 about 8° were 

 common. Images of moderately good resolu- 

 tion and intensity were obtained, but since 

 the surfaces were viewed obliquely they ap- 

 peared foreshortened. In such images the 

 magnifications are different in different direc- 

 tions and it is usual to call the maximum 

 magnification (in the direction perpendicular 

 to the plane of incidence) Wj. and the mini- 



mum magnification (in the foreshortened 

 direction parallel to the plane of incidence) 

 mi| . Then m|| = m^ sin 62 and with an 

 angle 62 of about 8° the foreshortening factor 

 (mj./w||) is about 7; hence there is a con- 

 siderable distortion of the image. This makes 

 it diflEicult to interpret the pictures — never- 

 theless, much work has been done with this 

 type of arrangement with useful results. 



The electrons scattered at the specimen 

 change velocity by different amounts, which 

 can be quite large. Consequently, chromatic 

 aberration in the objective lens becomes the 

 factor limiting the resolution of the reflection 

 electron microscope. The best resolution so 



o 



far obtained is about 300 A and this is prob- 

 ably somewhere near the limit attainable 

 unless some means of reducing the chromatic 

 aberration (e.g., a velocity filter or achro- 

 matic lens) can be developed. The figure of 



o 



300 A represents the resolution in the direc- 

 tion of maximum magnification. The resolu- 

 tion in the foreshortened direction is worse 

 by the foreshortening factor since a circle 

 of confusion at the image corresponds to an 

 ellipse at the specimen. 



It would obviously be desirable to reduce 

 the foreshortening and several methods of 

 doing this have been tried. The simplest is 

 to view the final image through a cylindrical 

 glass lens arranged to magnify it in the fore- 

 shortened direction. An alternative is to use 

 a cylindrical electron lens (Fert, 1956) to 

 correct the distortion before the image is 

 formed on the fluorescent screen of the elec- 

 tron microscope. This approach is not very 

 fruitful since the resolutions differ in differ- 

 ent directions in the image. The most promis- 

 ing method is to use larger values of 62 . The 

 main reason for the common use of small 

 values of 62 is that the intensity of the image 

 falls off very rapidly as the total angle of 

 deviation is increased. It is difficult to see to 

 focus the image if 62 is too big. However, by 

 using a suflEiciently powerful electron gun 

 Fert and his colleagues at Toulouse have ob- 

 tained good images with 61 = 2° and 62 = 



224 



