Phase diagrams for H/Ni(111) based on model interactions: Effects of strong long-range attractions

Abstract

We test the pairwise adatom-adatom interaction sets for H/Ni(111) computed by Muscat by using Monte Carlo techniques to determine the lattice gas phase diagrams they imply and then comparing with experiment. Although the correct ordered phase is reproduced and the maximum disordering temperature is approximately correct, there remain topological discrepancies with the experimental phase diagram. The origin of the main discrepancy is a relatively strong long-range attractive interaction. We describe the behavior this interaction causes and outline the circumstances under which similar phenomena might occur in other systems, developing insight through (re-)examination of the square-lattice Ising metamagnet problem. Finally we show how two previously studied interaction sets with relatively weak sixth-neighbor attractions lead to more accurate phase diagrams. One appendix collects some results on the metamagnet. Two others present transfer matrix scaling calculations of the critical properties of the disordering of two ordered states on a honeycomb lattice: hard hexagons (nearest neighbor exclusion) is Ising-like to high precision, while graphitic (2 × 2) is 4-state-Potts-like.