Reentrant first-order layering transitions in multilayer argon films on graphite

Abstract

We report ellipsometric coverage-vapor-pressure isotherm measurements for argon on graphite over the temperature range 65 to 84 K, which resolve up to eight layers. We confirm previous observations that the first-order layering transitions broaden near 67 K. However, in the fourth and higher layers, condensation again becomes sharp and apparently first order in the range between about 73 K and a new series of layer critical points near 77 K. Thus the surface-roughening transition is above 77 K and the behavior near 67 K represents disordering of only the top layer of the film.PACS numbers: 68.55.Jk, 64.70.Dv, 68.45.Gd A face of a crystal is said to exhibit surface melting l if a mobile "quasiliquid" layer exists on the equilibrium surface at a temperature below the bulk melting point T m , and its thickness increases with temperature, diverging as T approaches T m . Surface roughening 2,3 is the unlocking of the crystal surface from a particular lattice plane through the proliferation of steps, resulting in the loss of faceting in thermal equilibrium. 3 This transition is preceded on the low-temperature side by the appearance and proliferation of vacancies in and adatoms on the initially smooth surface plane. As the temperature increases towards the roughening temperature TR, the characteristic size of clusters of adatoms and vacancies diverges, and the clusters gain their own (divergent) adatom and vacancy clusters. Roughening is thus an essentially many-layer phenomenon, but it must be preceded by short-range disordering of the surface layer. Simulations show a strong correlation between disordering of the top layer and onset of a large surface mobility. 4 Therefore, while surface roughening and surface melting are quite distinct many-layer phenomena, their precursors in the top layer may be coupled.The properties of the surfaces of bulk crystals should be reflected in the properties of the surfaces of multilayer wetting films of the same solid on a foreign substrate. 5 " 7 Disordering of the (partially filled) top layer of a film is a transition of the 2D Ising class. A succession of Ising critical temperatures T Ct " is expected for films of different coverages near n-j layers, but always the transition is in the top layer. Using different models, Huse 8 and Nightingale, Saam, and Schick 9 have shown that the sequence T Cyn should converge to the roughening temperature, with TR -T c , n~-(lnn) ~2.Zhu and Dash 7 (ZD) have reported evidence for both surface melting and surface roughening in heat-capacity studies of two-layer to about twenty-layer films of both neon and argon on graphite. The evidence for surface roughening is small heat-capacity peaks near 0.8T m , attributed to disordering of the partially filled top layer. Previous volumetric isotherm measurements for argon on graphite by Gilquin 10 found broadening of the second-

Zippering and Intermeshing: Novel Phase Diagrams for Interfaces and Films

“…The transition occurs when thermal fluctuations overcome the local (e.g., nearest neighbor) interactions, but stronger, longer ranged (e.g., second neighbor) interactions exist to stabilize an overall flat interface [2]. At higher T the interface finally roughens completely.Recent experiments involving the growth of thin films on a substrate [4,5] have been interpreted as strong evidence for preroughening behavior [6]. The key feature is a sequence of reentrant first order layering transitions between integer-plus-one-half coverages lying above the usual low temperature layering transitions between integer coverages (see Fig.…”

Layering transitions, disordered flat phases, reconstruction, and roughening

“…In this last section we briefly compare the theoretical results we have obtained from the RSOS and sine-Gordon models with the results of the experiments on noble gases on graphite substrates [11][12][13][14]18 and discuss other possible interpretations of the data. We end by discussing work for the future.…”

“…Conversely, the bulk surface phases and phase transitions must be reflected somehow in the behavior of a finite but sufficiently thick film. Motivated by the results of some recent experiments on rare gases adsorbed on graphite [11][12][13][14] and MgO ͑Ref. 15͒ substrates, the purpose of this paper is to explore precisely this latter issue.…”

“…The discontinuities are, therefore, centered on integer coverages, as seen in the experiments. [11][12][13] The reentrant interval will shrink as h 2 increases. It will shrink as well as K increases to positive values, but grow as K decreases to negative values.…”

“…1͑c͒, except that the integer and half-integer layering lines are interchanged. Ellipsometry 11 and vapor pressure isotherm measurements, [12][13][14]18 however, are sufficiently accurate to rule out such an interchange. The continuation of path 4 shown in Fig.…”