Originally published in Ann. Fhtjs. 65, 216 (1921). 



10. SELF-DIFFUSION IN SOLID LEAD 



J. Groh and G. Hevesy 

 From the Chemical Institute of the School of Veterinary Medicine of Budapest 



We have recently shewni that the velocity of self-diffusion, that is, 

 the velocity with which the atoms (molecules) of molten lead change 

 places, can be ascertained by determining the velocity with which a 

 radioactive lead isotope spreads in molten lead. Experiments will now 

 be discussed whose purpose is the determination of the self-diffusion 

 velocity in solid lead. 



The extraordinarily high resistances which oppose place exchange in 

 the solid state led from the outset to the expectation of very slow self- 

 diffusion in soUd lead ; we have, therefore, avoided setting up experiments 

 at room temperature and have sought rather to determine the self- 

 diffusion in lead heated and maintained about 40 °C below its melting 



point. 



Several series of experiments lasting from 1 to 3 months showed that 

 the self-diffusion velocity of lead at 280°C, that is, 46° below its melting 

 point, is less than 0.001 cm^/day. A series of experiments was then 

 performed in which lead filaments, about 2 cm long, were heated for 

 more than 400 days ; these filaments consisted, as will be described 

 in detail below, of a 1 .5 cm long inactive and a 0.5 cm long active portion 

 of lead. No diffusion of the active lead isotope into the inactive lead 

 could be detected even after this long period of experiment. The self- 

 diffusion constant of the solid lead is accordingly still smaller than 

 0.0001 cm2/day, even at a temperature of 280°, since values of this order 

 could still have been easily determined in the stated conditions. 



This result is not without interest, especially when it is compared 

 with the well-known Roberts-Austen experiments^. Roberts-Austen 

 allowed gold to diffuse into solid lead and found the diffusion constants 

 recorded in the table below, which also includes our experimental result. 



Even at 251°C, therefore, the diffusion of gold into lead is at least 

 three thousand times as fast as that of lead into lead at 280°C, the 



ij. Groh and G. Hevesy, Ann. Phys. 63, 85 (1920). 

 2 W. C. Roberts-Austen, Phil. Trans. 187, 404 (1896). 



