ACTIVATION OF ELECTRICAL CONTACTS BY ORGANIC VAPORS 771 



ation is much complicated by burning of carbon and by the impedance 

 offered by air to diffusion of molecules to the electrode surfaces. Because 

 of these complicating factors, activation will not occur in air if the vapor 

 pressure is too low or if the time between arcs is too short. 



Arcs at the making and breaking of clean contacts — clean in the sense 

 that they are free from carbon — produce transfer of metal from one con- 

 tact to the other with a resulting pit and mound of about equal volumes. 

 The situation is greatly changed by carbon. The presence of carbon 

 causes increased arcing, alters the characteristics of the arcs, and greatly 

 changes the resulting erosion both in character and amount. With carbon 

 present, some or even all of the eroded metal does not stick to the elec- 

 trodes, and there is often loss of metal from both of them, the missing 

 metal turning up mixed with carbon in a loose black powder. With car- 

 bon on the surfaces, successive arcs occur at different places, and the 

 resulting erosion tends to be smooth with the electrodes worn down uni- 

 formly all over their surfaces. This is because each arc burns off carbon 

 at its center, while it produces more around its periphery where the 

 metal is cooler, and each new arc strikes on a newly carbonized surface. 



Everj^ arc, of either the active sort or of the "inactive" sort which 

 occurs at clean surfaces, is predominantly an arc in metal vapor. The 

 active arcs, as well as the arcs at clean surfaces, are of one or the other 

 of two quite distinct types which have been called, respectively, "anode 

 arcs" and "cathode arcs" (Reference 3 and 4 which are concerned with 

 palladium electrodes only). In an anode arc, most of the metal of the 

 arc is vaporized from the anode by electron bombardment, but in a 

 cathode arc the metal is supplied from the cathode by the explosion of 

 small areas due to Joule heating by field emission currents of enormous 

 densities flowing through them. In an anode arc, the erosion is predom- 

 inantly from the anode, and in a cathode arc from the cathode. 



Whether a particular arc is of the anode or of the cathode type is de- 

 termined by the electrode separation and the contact metal. For pal- 

 ladium electrodes, an arc is an anode arc if the separation is less than 

 about 0.5 X 10~^ cm, but a cathode arc if the separation is greater than 

 this \-alue. The corresponding critical distance for silver is 3 or 4 X 10~* 

 cm. The carbon particles producing activation permit breakdown at 

 separations for which it would not occur in the absence of carbon, and 

 thus favor cathode arcs. The critical distance of palladium is so small 

 that all active palladium arcs are cathode arcs, with the greater loss of 

 metal from the cathode. For silver, on the other hand, the critical dis- 

 tance is so large that active arcs at silver surfaces are in many cases anode 

 arcs, with the greater loss of metal from the anode as in the case of inac- 

 tive silver arcs. 



