Abstract
Introduction
Experimental Setup
Data
Discussion
Results
References
|
 |
Introduction
 At the triple point of a one component
system, bulk solid, liquid, and vapor can coexist, but even a small thermodynamic
perturbation will break this degeneracy . The interfaces at the substrate
and the vapor of an adsorbed film provide such a perturbation. For this
reason, the phase of the film is not uniquely determined by the bulk phase
diagram. The consequences of interfacial effects are surprisingly subtle
and determine phenomena such as non wetting of solid films [1,2], surface
melting [3], nucleation, and triple point induced wetting [4-8] and dewetting
[9]. Pandit and Fisher [10] have discussed a number of possible surface
phase diagrams. They show that typically surface phases are separated by
first order phase transitions at temperatures distinct from the bulk triple
temperature, Tt. Nevertheless, experiments near solid-liquid-vapor
triple points have never revealed transitions except at Tt.
Previous experiments have relied
on two types of techniques. The first utilizes thermodynamic measurements
on exfoliated graphite substrates [11]. Recently it has become clear that
multilayer adsorption on these substrates is always accompanied by capillary
condensation, so it is very difficult to determine surface phases near
coexistence [12]. Another class of experiments, including the one reported
here, utilize microbalances [5-9, 13] which do not suffer from this complication.
All previous microbalance experiments have identified a single particularly
simple type of triple point wetting scenario in which a thin liquid-like
film is stabilized on the substrate below Tt, and the thickness
d diverges as a power law d~(-t)n, with t=(T-Tt)/Tt.
The exponent n is approximately -1/3, which reflects the fact that the
chemical potential of the liquid below Tt is larger than the
chemical potential at solid-vapor coexistence by an amount proportional
to t. This must be compensated by the decrease in the chemical potential
due to the van der Waals interaction with the substrate, which is proportional
to 1/d3. The experiments we describe below provide evidence
for a distinctly different type of surface phase behavior. In particular,
for Ar on gold, we find that near bulk coexistence, a layered solid/liquid
film structure is stabilized in an approximately 3K interval which brackets
Tt. The layered phase melts for temperatures both above and
(surprisingly) below this interval. The thickness of the low temperature
liquid film obeys a power law in -t. For methane on gold, films below Tt
are always solid, and the thickness is almost independent of t.
|
|
