USE or FIELD EMISSION ELECTRON MICROSCOPE 927 



Migration of Ba on W 



Starting at about 800°K, Ba migrates over large distances — from one 

 side of the paraboloid to the other or about 4000 A. Figure 9 shows the steps 

 in the migration process and the early stages of evaporation. The migration 

 process can be followed continuously by observing the screen for moderate 

 V between 800 and 1000°K. Migration is essentially complete after five 

 minutes at 1045°K. By then the crystallites have disappeared completely 

 and with them abnormally high local fields. It is therefore possible to 

 compute the field from the appHed voltage. Then, as explained above, 

 values of (p and 6 averaged for the whole surface can be computed. In this 

 way we find that for photos c and d in Fig. 9, <p = 2.00 and 6 is about 1.00. 



Evaporation of Ba on W 



For T > 1200° evaporation can be observed in five minutes. This is evi- 

 denced by the fact that after such glowing the value of V required for a 

 given current increases. The details of the evaporation are continued in 

 Fig. 10. From the values of V and i, values of (p and 6 have been calculated 

 and are shown in Table I. From this table it appears that nearly all the Ba 

 is evaporated in five minutes at 1600°K. 



Further information on how the evaporation rate at a given T varies with 

 d can be deduced from experiments for which no photos are shown. Suppose, 

 in the above series of experiments, the point had been glowed for twenty 

 minutes instead of five minutes, the calculated Ba concentration 6 would 

 have reached a somewhat lower value than .80, say .75. Still further glowing 

 would have reduced only sHghtly. From this we conclude that at T = 

 1200 and ^ = .75 the rate of evaporation or dS/dt is so small that additional 

 glowing for twenty minutes reduces 6 by small amounts. If now T is raised 

 to 1300°K for one minute, 6 is substantially reduced, perhaps to .65. After 

 five minutes at 1300°K, 6 might be .55. After twenty minutes at 1300, 

 might be .51. Long times at 1300°K might reduce to .50. Only by raising 

 T above 1300°K could be substantially reduced below .50. These observa- 

 tions suggest that the rate of evaporation of Ba on W depends not only on 

 T but also on 0: for a constant T it is substantially for all values of less 

 than a critical value 0c. Above 0c, the evaporation rate increases rapidly 

 with 0, perhaps exponentiaUy. Hence the probability of evaporation of a 

 particular Ba atom depends on the proximity of neighboring atoms. This 

 must mean that the forces between adsorbed Ba atoms are comparable to 

 though smaller than the forces between Ba and W. A plot of the values in 

 Table I vs T would show that 0c varies linearly with T between 1130 and 

 1430 or between = 1.00 and .18. 



