| ||
  | ||
  | ||
Note:
1. The data of sizes 219, 477, 879 and bulk are from F. Ercolessi, W. Andreoni, and E. Tosatti, Phys. Rev. Lett., 66, 911 (1991). Others are from our simulation of gold heating.
2. The data from gold heating are read directly from the plots of radii of gyration and bond order paramters, so they may not be very accurate.
3. The melting temperature measured from continuous heating may vary
a lot for small system (roughly less than 1000) with different initial
structures. For example, the melting temperature of gold sphere of 603
atoms with FCC initial structure is about 670K, but that with icosahedral
initial structure is about 980K.
From H.B.Liu et al. Surf. Sci. 491 (2001) 88-98
Note: Tight-binding model gives different temperature scale from glue
model.