Superfluid Onset and Prewetting of 4He on Rubidium

J. A. Phillips, D. Ross, P. Taborek and J. E. Rutledge 
Department of Physics and Astronomy 
University of California, Irvine 92697 

(1998)
 
 
Abstract 

Introduction 

Experimental Methods 

Data 

Discussion 

Conlcusions 

References 
 
 
 
 

Projects Page 
 
 
 
 

Introduction

Films of 4He adsorbed on solid surfaces have been studied for many years and a standard picture of their properties on most surfaces has been established [1]. The interaction between a 4He film and nearly all surfaces is so strongly attractive that the first one or two atomic layers of the film are solidified or otherwise localized. Subsequent layers remain fluid and become superfluid at low temperature. The experimentally observed signatures of superfluid onset are remarkably independent of the properties of the substrate; the 2D superfluid transition has the same characteristics whether it occurs on solid hydrogen, mylar or gold. Similarly, the standard Kosterlitz-Thouless (KT) model [2,3] used to describe the superfluid transition in films is strictly two dimensional and is independent of the boundary conditions at the substrate. The KT picture has been extremely successful in quantitatively explaining the shear response [4], thermal conductivity [5] and third sound modes [6] of 4He films near superfluid onset on a variety of substrates.
Recently, alkali metal surfaces have proven to be radically different adsorption substrates for 4He and other simple gases. Adsorption potentials for 4He on the alkali metals are “weak” which means that the potential well formed by the competition between long-range van der Waals attraction and short range Pauli repulsion has a depth comparable with the binding energy of a 4He atom in the bulk fluid. The phases of helium on alkali metal substrates have been theoretically analyzed by Cheng, Cole, Saam and Treiner [7]. Their analysis has been confirmed by experiments on the weakest substrate, cesium. Cs is non-wet at low temperatures and wet above a wetting temperature near 2 K. Between 2 K and about 2.5 K it displays a prewetting transition [8], a first order phase transition on the vapor side of the bulk liquid-vapor coexistence curve marked by two-phase equilibrium between thin and thick films. The effects of first order prewetting on the KT transition have also been investigated [9]. In the case of thin cesium substrates, the KT line terminates on the prewetting phase boundary at a critical endpoint. For temperatures above the critical endpoint, prewetting and the KT transition are distinctly different transitions. Below the critical endpoint, the KT transition does not exist and prewetting is a direct transition between a thin normal phase and a thick superfluid phase. Since the thin phase is always normal on Cs, superfluid onset can only be studied above the wetting temperature. 
There have been several previous  investigations of 4He adsorption on Rb . Rubidium is a slightly stronger substrate than cesium, and is theoretically expected to have a lower wetting temperature [7,10]. A lower wetting temperature provides an experimental opportunity to study superfluid onset over a wider temperature range. Superfluid onset on weak substrates is particularly interesting because even the first adsorbed 4He layer is expected to remain fluid so the inert layer of solidified helium which forms on strong substrates can be avoided [7]. 
There have also been experimental investigations of 4He adsorption on Rb [11-14].  The most extensive of these studies have utilized heat flow measurements to map the superfluid onset. This type of experiment can detect transitions to a superfluid state, but, as shown in [9], superfluid onset and prewetting are not necessarily coincident phase transitions. In this paper, we report experimental observations of both the wetting and superfluid properties of 4He films on Rb using the quartz microbalance technique. Characteristic features in the frequency shift and the dissipation allow us to construct a phase diagram showing the boundaries between thick and thin and superfluid and normal phases. Above 1.9K, prewetting and superfluid onset are distinct phase transitions, while below this temperature, the transitions are strongly coupled. At temperatures below 1.9K, superfluid onset on Rb is surprisingly different from the Cs case investigated in reference [9]. Although the superfluid transition on Rb retains Kosterlitz-Thouless type features, the transition is hysteretic and the size and temperature dependence of the features do not conform to the universal predictions of the KT theory. Another important practical difference between Rb and Cs is that our technique for producing alkali metal substrates is less reproducible for Rb. Isotherms on Rb films evaporated and annealed in nominally the same way had features whose location was reasonably well defined, but whose height, width and shape varied substantially from one substrate to another. 
This paper is organized as follows. Section II contains a discussion of the experimental procedures, including the techniques used for preparing the Rb substrates and a description of a method for determining the chemical potential at low temperatures when the vapor pressure is too low to measure directly. In section III, we present the data, which consist primarily of measurements of the frequency shift and the dissipation of the quartz microbalance as a function of 4He chemical potential at constant T. A number of these isotherms from 0.15K to 2.2K are used to assemble a m-T phase diagram that shows the relationship between the prewetting line and the KT transition line [15]. Measurements of the superfluid density and dissipation at onset show that in many cases, superfluid transitions on Rb do not follow the standard KT model. Evidence of two types of behavior in the thin film phase above and below 0.3K is also presented. In section IV we discuss the wetting and superfluid onset behavior of 4He on Rb and compare it to results obtained with Cs and conventional strong substrates such as Au. The possibility that the KT line meets the prewetting line in a multicritical point and the possible role of substrate disorder is also considered. Finally, section V provides a summary of our findings.