SCANNING 



the interpretation of the geometry of the 

 image is not difficult , particularly if stereo- 

 micrographs are taken. Angles between 20° 

 and 45° are commonly employed, giving fore- 

 shortening ratios of between 3:1 and 1.4:1, 

 respectively. The position of the collector 

 does not affect the geometry of the image. 



Contrast in the image depends upon sev- 

 eral factors, chief among which are the ener- 

 gies of the primary and secondary electrons, 

 the observation angle of the specimen and 

 the arrangement for collection of the emitted 

 electrons. 



Accordhig to their energies, electrons 

 which are emitted from the surface of the 

 specimen may be classified into two groups. 

 There are the true secondary electrons which 

 have a mean energy of about 6 eV and a 

 maximum of about 40 eV, and there are the 

 electrons mostly with energies of between 

 one-half and three-quarters of the primary 

 beam energy which may be termed "re- 

 flected" electrons. 



The collector may be made to respond to 

 either or both of these groups, the resulting 

 image being characteristic of the group de- 

 tected. Collection of the secondary compo- 

 nent has been found to give the most infor- 

 mative image and is the normal mode of 

 operation in the scanning microscope. Con- 

 trasts in this case result mainly from changes 

 in the local angle of incidence. If only the 

 reflected component is collected, contrast 

 is to a certain extent dependent also upon 

 the surface composition. For example, there 

 is an appreciable difference between brass 

 and aluminum in the magnitude of the re- 

 flected component (17). 



If only the secondary component is col- 

 lected, then the signal strength, s, at the 

 output of photomultiplier may be repre- 

 sented quite closely by 



s = K cosec e for 20° < < 40° 



where K h a, constant 

 thus 



Ss = —K cosec e cot 686 



5s /s 



■cot 686 



(13) 



Taking the threshold contrast as 5%, i.e., 

 8s /s = 0.05, equation shows that a change 

 in angle of 1.3° at 6 = 25° should be detect- 

 able. This of course applies only to small 

 changes in the angle between relatively 

 large flat areas. For rough surfaces the 

 contrast mechanism is somewhat different. 

 Consider a large asperity BCD, on the 

 surface of the specimen (Figure 4). The re- 

 gion ABC is partially screened from the 

 collector with the result that the electron 

 current entering the collector from this 

 region will be reduced. However, a propor- 

 tion of the secondary electrons will still be 

 collected since, if emitted at a favorable 

 angle, these electrons may follow curved 

 paths to reach the collector, the input of the 

 collector being normally biased a few hun- 

 dred volts positive. Thus detail may be seen 

 in the image from such regions if the second- 

 ary component is collected, whereas in the 

 case of the reflected component, no detail 

 may be seen because these electrons, having 

 high energy, can follow only straight -line 

 paths to the collector. 



At the tip of the asperity, C, there will be 

 a region thin enough to allow complete pene- 

 tration of the primary beam and a large 

 number of secondaries will be released on the 

 side of the asperity nearest the collector. 

 This will give rise to a region of high inten- 

 sity in the image. At beam voltages of 10-20 

 kV, the mean depth of penetration into a 

 metal is of the order of 1-2 fj. and the bright 

 bands along thin edges in the scanning 



DIRECTION OF 

 INCIDENT BEAM 

 AND DIRECTION 

 OF OBSERVATION 



COLLECTOR 



Fig. 4. Image formation at a large asperity, 



247 



