ELECTRON MICROSCOPY 



t 200-500 V 



+ 6-IOkV 



ALUMINIZED 

 PLASTIC 

 SCINTILLATO 



input/ 



GRID 



LIGHT 



PIPE |l,PHOTO- 



Jbcathode 



INSULATION 



METAL TUBE 



Fig. 5. Electron collector using a plastic scin- 

 tillator. 



image are of the same order. Under these 

 conditions smiace detail on the mcident side 

 of a thin edge can have httle effect on the 

 magnitude of the collected current and in 

 fact no detail can be seen m such regions of 

 high intensitj'. The beam accelerating volt- 

 age in the scanning microscope is made rela- 

 tively low in order to reduce the penetration 

 and, therefore, the regions of high intensit3\ 

 On parts of the smiace remote from thin 

 edges, the secondaries released b}' deeply 

 penetrating primaries cannot reach and es- 

 cape from the surface. Wells (8, 17) has 

 shown theoretically that the diameter of 

 the region from which the secondaries escape 

 does not exceed the diameter of the inci- 

 dent beam by much more than 100 A. Under 

 these conditions, penetration effects are 

 onlj' hkely to be important when resolution 

 of detail of the order of 100 A, or less, is 

 attempted. Resolutions of about 300 A have 

 so far been obtained with the scanning mi- 

 croscope. 



The Collector and Amplifying System. 

 The current leaving the specimen is of the 

 order of 10"^^; a simple electrode collector 

 followed by an amplifier would introduce 

 too much noise at this low level, conse- 

 quently, some form of electron multipli- 

 cation is necessary. In ]\Ic]\Iullan's scan- 

 ning instrument, a direct electron multiplier 

 was used which, while giving very satisfac- 

 tory resuks, necessitated a somewhat com- 



pHcated floating head-amplifier and intro- 

 duced various other complications. 



A collection system devised by Everhart 

 (9, 17) avoids many of the difficulties of the 

 earlier S3\stems. The secondaries are acceler- 

 ated towards a copper mesh grid covering 

 the end of a short metal tube (Figure 5). 

 The tube and grid are at a potential of a 

 few hundred volts positive with respect to 

 the specimen. After passing through the 

 grid, the electrons are further accelerated 

 through a potential of several kilovolts to 

 strike an aluminized plastic scintillator. 

 Each electron produces a large number of 

 photons which are registered by a photomul- 

 tiplier via the light pipe. In this system, 

 each collected electron produces several 

 electrons from the photo-cathode of the 

 multiplier, hence, the conversion is noise- 

 free. The scintillator has a much shorter 

 time constant than a fluorescent screen at 

 low intensities and is adequate to deal with 

 the highest video-frequencies used. This ar- 

 rangement also has the additional advantage 

 of allowing direct coupling throughout the 

 system, thus avoiding the necessity for any 

 form of beam modulation. 



A band width of about 160 kc/s is used in 

 the video amplifier. This passes satisfac- 

 torily the highest ^-ideo-frequencies which 

 occur when viewing the picture directly. A 

 gamma-control stage (20) for controlling 

 contrast and correcting nonlinearity in the 

 display tube has been found to be a useful 

 feature. 



The Time Bases, Display and Record- 

 ing Systems. The time bases provide a wide 

 range of sweep speeds for direct obsen-ation 

 of the picture and for recordhig. During 

 direct obser\'ation, the picture is displayed 

 at the rate of about one frame per second 

 with a maximum of about 400 lines per 

 frame. By using a long persistence screen on 

 the display tube, three or four successive 

 frames are integrated, thus improving the 

 signal-to-noise ratio considerably. 



For recording, a range of sweep speeds 



248 



