802 THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1957 



trode surfaces, with heat producing decomposition into carbon and 

 hydrogen rather than evaporation of undamaged molecules. Surface ' 

 films prevent such strong adsorption, and metals with surfaces that are 

 normally covered by oxide films cannot be activated. 



In some very recent experiments in extremely high vacuum, P. Kis- 

 liuk has found^'' that benzene molecules are strongly adsorbed upon a 

 tungsten surface that is perfectly clean, but if there is on the surface just 

 one single layer of oxygen molecules, benzene molecules are not ad- 

 sorbed. M. M. Atalla has reported (Reference 5, page 1090), on the 

 other hand, that tungsten (and nickel also) can be activated if the pres- 

 sure of air is as low as 10~* mm Hg. It seems probable that arcs at op- 

 erating contacts remove adsorbed oxygen temporarily, and at sufficiently 

 low air pressures this may be replaced in part by organic molecules rather 

 than by oxygen. ^ 



Even at palladium surfaces, some cleaning by arcs seems to be neces- 

 sary before benzene molecules can be adsorbed. This conclusion is reached 

 in unpublished adsorption experiments carried out by W. S. Boyle upon 

 palladium surfaces in air containing benzene vapor. In this work, two 

 optically flat palladium surfaces are separated by an exceedingly small 

 distance to make an electrical capacitor. With a very sensitive capaci- 

 tance bridge, one can detect the change in capacity that would be 

 produced by the adsorption on the palladium surfaces of even a small 

 fraction of a monolayer of benzene molecules. In experiments carried 

 out with this equipment it was found that benzene molecules are not 

 adsorbed upon a palladium surface in air at atmospheric pressure. To 

 reconcile this conclusion with the well known facts of activation, it seems 

 necessary to conclude that even a palladium surface can adsorb benzene 

 molecules only after it has been partly cleaned by arcing. 



5.3 Alloys 



When a base metal is mixed with a noble metal, the result can be an 

 alloy which is activated less readily by organic vapors than would be 

 the noble metal constituent alone. In the curve of Fig. 11 is plotted the 

 number of operations required under a particular set of standard condi- 

 tions to activate a series of alloys of palladium and nickel. In air, nickel 

 itself cannot be activated at all. The amount of carbon formed from 

 benzene decomposition on the surface of a palladium-nickel alloy is 

 always less than the amount which would be formed under the same 

 conditions upon pure palladium. One does not know whether benzene 

 is held less firmly on the alloy sui'face so that there is more likelihood 



i 



