arXiv subjects

Douglas Davidson

Publications and source records attributed to Douglas Davidson.

3 recordsLinked to original sources

arXiv1999

Facet ridge end points in crystal shapes

Equilibrium crystal shapes (ECS) near facet ridge end points (FRE) are generically complex. We study the body-centered solid-on-solid model on a square lattice with an enhanced uniaxial interaction range to test the stability of the so-called stochastic FRE point where the model maps exactly onto one dimensional Kardar-Parisi-Zhang type growth and the local ECS is simple. The latter is unstable. The generic ECS contains first-order ridges extending into the rounded part of the ECS, where two rough orientations coexist and first-order faceted to rough boundaries terminating in Pokrovsky-Talapov type end points.

cond-mat.stat-mech↗

arXiv1998

Temperature Dependence of Facet Ridges in Crystal Surfaces

The equilibrium crystal shape of a body-centered solid-on-solid (BCSOS) model on a honeycomb lattice is studied numerically. We focus on the facet ridge endpoints (FRE). These points are equivalent to one dimensional KPZ-type growth in the exactly soluble square lattice BCSOS model. In our more general context the transfer matrix is not stochastic at the FRE points, and a more complex structure develops. We observe ridge lines sticking into the rough phase where thesurface orientation jumps inside the rounded part of the crystal. Moreover, the rough-to-faceted edges become first-order with a jump in surface orientation, between the FRE point and Pokrovsky-Talapov (PT) type critical endpoints. The latter display anisotropic scaling with exponent $z=3$ instead of familiar PT value $z=2$.

cond-mat.stat-mech↗

arXiv1996

Roughening Induced Deconstruction in (100) Facets of CsCl Type Crystals

The staggered 6-vertex model describes the competition between surface roughening and reconstruction in (100) facets of CsCl type crystals. Its phase diagram does not have the expected generic structure, due to the presence of a fully-packed loop-gas line. We prove that the reconstruction and roughening transitions cannot cross nor merge with this loop-gas line if these degrees of freedom interact weakly. However, our numerical finite size scaling analysis shows that the two critical lines merge along the loop-gas line, with strong coupling scaling properties. The central charge is much larger than 1.5 and roughening takes place at a surface roughness much larger than the conventional universal value. It seems that additional fluctuations become critical simultaneously.

cond-mat.stat-mech↗