Reconstruction at Surfaces

Plausibility Argument for Surface Relaxation. As already noted above, crystal surfaces are characterized by a wide range of structural rearrangements relative to the ideal surface. One such rearrangement is surface relaxation, with the entire surface layer moving either towards or away from the subsurface layers as shown 

Carlsson (1990)). Double and triple lines associated with surface atoms are meant to indicate the strengthening of bonds associated with those atoms that have lost neighbors due to the free surface. 

We have already emphasized the revolution in surface science ushered in on the heels of high-quality vacuum technology and surface scanning probes. One subbranch of surface science that has been the beneficiary of these advances is the study of metal surfaces. The intention of this subsection is to consider two distinct case studies, namely, the surfaces of fee Au and those of bcc W. These choices are inspired by the combination of high-quality experimental data, richness of surface phases, their diversity from the standpoint of the demands they place on atomistic simulation and their relevance to later generalizations to surfaces with steps, islands and even more extreme forms of roughness. 

The experimental data associated with this surface are actually more revealing than we have suggested (Gibbs et al. 1990, Abernathy et al. 1992). On the basis of systematic X-ray reflectivity studies, it has been deduced that a sequence of phases are found as a function of temperature these phases are shown schematically in fig. 9.12. In particular, two related pseudohexagonal phases have been foimd which differ by a small relative rotation in addition to a high-temperature disordered phase in which the pseudohexagonal symmetry is lost. However, despite the apparent abruptness and reversibility of the transition between the two distorted hexagonal phases, it remains unclear whether or not these structures constitute true equilibrium phases. 

The mission of any energetic methods that might be brought to bear on the structure of this surface is to explicate the origins of its tendency towards close-packing, even at the cost of abandoning its epitaxial relationship with the subsurface layers. Our strategy is to contrast pair fimctional and first-principles approaches to this reconstruction. In addition, we will examine the way in which first-principles calculations may be used to construct an effective Hamiltonian for 

---