Figure 1.  Heat Capacity of  a binary Kr / CCl4 film.  The signals at 95 and 107 K correspond to melting of the 3rd and 2nd layers, while the larger peak at 115 K signifies displacement of the Kr film by CCl4.

Figure 2: Isotherm data for Kr at 90 K.  Steps in the coverage represent formation of successive monolayers of film

Unique and rich "layer-by-layer" phase diagrams arise for these condensed films in the presence of the smooth, attractive graphite substrate.  Phase transitions such as layer condensation, melting, and commensurability changes are observed in individual layers, as are "2D" critical and triple points.  We hope to observe the evolution of the growing film as it thickens and begins to resemble the more familiar "3D" or bulk condensed matter.  For example, at some film thickness, we expect to see the end of the layer-by-layer continuous melting seen in films of several layers, and the onset of the first order melting observed in bulk solids.  Additionally, as the film thickens, the outermost layers become like a free surface of a bulk crystal or liquid.  Because of their "thinness", thin films can reveal much about surface physics in thermodynamic measurements, and we have seen evidence, in films as thin as three layers, of smooth, rough, and "pre-roughened" surface phases, the latter being a theoretically predicted surface state that is macroscopically flat, but microscopically disordered. 

The experimental workhorse for these studies has been the "Adiabatic Scanning Ratio Calorimeter" designed and constructed by Mark Lysek and Peter Day.  The apparatus features a 100 cc cell of exfoliated graphite foam as an adsorption substrate with almost 600 m² of surface area, and permits isotherm measurements in addition to precision calorimetry on a film over a wide temperature range.  Studies of, for example, CH₄ and Kr, reveal many of  the phases described above and the rich phase diagrams shown below. 
 
 

Both systems show layer-by-layer melting occuring well above the bulk triple points.  We believe the "zig-zag" feature in the Kr phase diagram, which connects to the 3rd layer melting line, signifies the thin film limit of the pre-roughening transition for a bulk solid, and that we are thus seeing the onset of bulk film behavior.  However, formation of  third and higher layers is accompanied by capillary condensation in the pores of our graphite foam, hindering and obscuring further film growth.  Thus, the exciting prospect of following the observed melting transitions and proposed pre-roughening transition into the bulk is not possible in the present system. 

Our current efforts focus on extending the measured phase diagrams into higher layers of the film by somehow avoiding the problem of capillary condensate.  One technique involves binary film systems, where a preadsorbed layer of some highly condensable material, like carbon tetrachloride, is displaced off the graphite surface by an inert gas film, like Kr.  The hope is to fill the pores with the preadsorbed CCl₄ and thus prevent the Kr from ever capillary condensing.  While this technique has not as yet led to an understanding of thicker films, the physics of the binary film systems has proved interesting in itself, and we are investigating new phase transitions unique to the two component films.  Additionally, we seek new graphite substrates, in flat geometries without the pores that allow capillary condensate.  We are exploring the possibility of plating graphite onto a quartz crystal or microscale membrane, and thus study thin films with a microbalance technique.