Substrate Effects in Superfluid Helium

The self emptying beaker technique was used to study the superfluid properties of ^3He confined in the van der Waals film adsorbed on the surface of a metal beaker. The experimental cell was designed to minimize thermal gradients along the ^3 He film. In contrast to the results of an earlier experiment by Sachrajda et al, which suggested that film flow occurred at temperatures as high as 3.5 mK (SACH-85), no flow was observed above the bulk transition temperature T_sp{rm c}{rm B} = 0.93 mK. The transition temperatures measured using round rim beakers agreed with theory, giving the predicted normal-superfluid phase boundary 2 delta/xi(T) = pi, where delta is the film thickness and xi(T) is the temperature dependent coherence length. The ^3He film thickness was inferred from Atkins' oscillation measurements of ^4He films on the same substrate. When a ^4He monolayer was adsorbed on the surface of a copper beaker, it suppressed the diffuse scattering of ^3He quasiparticles at the copper wall, an effect first observed by Freeman et al using a mylar substrate (FRMN-88). With the ^4He monolayer in place, there was no measurable suppression of the transition temperature, even for films as thin as 100 nm. This suggests that the ^3 He quasiparticle scattering at the free liquid surface as well as the ^4He covered substrate was specular. This is the first evidence of the nature of the scattering at the free surface. After the ^3He level in the beaker had dropped between 15 and 85 mu m, the flow rate abruptly dropped by a factor to ten or more. This may be associated with the transition between the superfluid B-phase, expected in thick films, and the superfluid A-phase, expected in thin films. The observed critical currents are roughly an order of magnitude smaller than predicted by the pair breaking limit, suggesting some other dissipation mechanism is responsible for limiting the current.