EVAPORATION OF ATOMS 381 



Values of Ti and Di (for Cs atoms in the ist layer) are given for 

 several temperatures in the 3rd and 2nd columns of Table VI. 



TABLE VI 



55uRFACE Diffusion Coefficients for Cs Adatoms in ist and 2nd Layers 

 Ti and T2 are the "lives" in elementary spaces in ist and 2nd layers as given by 

 Eqs. (63) and (64) ; x is the "evaporation life" in 2nd layer given by Eq. (49). 



The mobility of adatoms in the 2nd layer must be much greater than 

 that in the ist layer since the atoms in the 2nd layer are held by much 

 weaker forces. For low values of 61 at which Di was measured, the heat 

 of evaporation of atoms from the ist layer is 2.83 volts. The potential 

 barrier corresponding to the coefficient of i/T in Eq.' (63), 3082, is, as we 

 have seen, 0.61 volt which is 21.5 percent of the heat of evaporation. The 

 potential barrier separating the elementary spaces for the 2nd layer must 

 be much less than 0.61 volt which is 78 percent of the heat of evaporation 

 (0.79 volt) from the 2nd layer. It seems reasonable to assume that the 

 barrier in this case also is approximately 21.5 percent of the heat of 

 evaporation. This would give 0.17 volt. The Cs atoms iti the ist layer, 

 however, because of their larger size compared to W atoms, constitttte a 

 rougher support for the atoms in the 2nd layer than is provided for the 

 1st layer of atoms by the underlying tungsten surface. Thus we may adopt 

 the rough value 0.2 volt as most probable value for the barrier in the 2nd 

 layer. This corresponds to a coefficient lOOO for i/T. This gives for the 

 life T2 of an adatom in an elementary space of the 2nd layer 



logic T2 = — 15.09 + looo/r. (64) 



We have taken the term — 15.09 to be the same as in Eq. (63), since for 

 evaporation ^^ and for diffusion ^'^ this term remains nearly constant even 

 for different substances. 



Thus a general equation for the evaporation life t was found ^^ to be 



r = ^.7XlO-'-'Mh,T-'e^"'. (65) 



Because of the similarity of the processes of evaporation and of mobility, 



by which atoms hop from one position to another, we may expect this 



^® Reference 2. See particularly Eq. (37) on page 2806. 



*' S. Dushman and L Langmuir, Phys. Rev., zo, 113 (1922). 



