348 BELL SYSTEM TECHNICAL JOURNAL 



obtained by specifying a pellet containing 4.0 to 4.2 milligrams of 

 caesium chromate, an oxygen-caesium ratio of 19, and a heat treatment 

 at 220° C. for 25 to 30 minutes. It should be noted, of course, that 

 the above are initial specifications. Two thirds of the cjesium goes 

 to form the matrix and active film of the cathode surface, while the 

 remainder is deposited upon the inner surface of the glass bulb. 



Studies of the Cesium Thin Film 

 These cells show small fluctuations in sensitivity and shifts in 

 spectral response with time and with the temperature of the sur- 

 roundings. Even cells prepared under the optimum conditions take 

 several days to stabilize. The integral sensitivity often changes by 5 

 or 10 per cent, usually to higher values. But further changes occur 

 with changes in temperature. A number of cells aged at 50° C. 

 increased in sensitivity by about 10 per cent in a few days, and then 

 fell to their former stable values after operation for a few days at 

 room temperature. In particular, cells made with large ratios of 

 caesium to oxygen and short heat treatments, which should be con- 

 ducive to the presence of small residual amounts of free caesium, tend 

 to be the least stable, and to decrease rather than gain in integral 

 sensitivity. Such cells are apt also to have low insulation resistance 

 which tends to be unstable in value as compared to the normal cell 

 which has a high and relatively stable insulation resistance. 



Since the changes in sensitivity at 50° C. or below are generally 

 reversible, they can not be due to chemical reactions of the irreversible 

 type by which the cells are prepared. These changes seem most 

 easily explicable as due to changes in thickness of a thin film of free 

 caesium reversibly adsorbed upon the matrix of gross material, many 

 molecules thick, which has been formed upon the silver cathode 

 surface. According to this view it should be possible to correlate the 

 observed fluctuations in sensitivity with variations in film thickness 

 as the free caesium diffuses in and out of the underlying matrix, or 

 evaporates and recondenses upon the various surfaces of the photo- 

 electric cell interior. 



Now it is impossible to differentiate between caesium adsorbed on 

 the surface, absorbed in the matrix, or in chemical combination, 

 by the type of quantitative data considered in the previous section, 

 for all are placed simultaneously in their appropriate places and in 

 proper amounts by the processes described. Neither is it possible to 

 build up a matrix alone and subsequently place the film upon it with 

 any assurance that the final state shall be comparable to that found 

 in the cells already described. So resort has been had to a variational 

 method of studying the surface film. 



J 



