Quantum films adsorbed on graphite: Third and fourth helium layers

Abstract

Using a path-integral Monte Carlo method for simulating superfluid quantum films, we investigate helium layers adsorbed on a substrate consisting of graphite plus two solid helium layers. The solid helium layers are modeled first as inert, with paths frozen at equilibrated positions, and then as active, with second layer atoms included in the Monte Carlo updating. In both cases, we observe the formation of as many as three well defined additional layers above the first two, and determine the layer promotion density by calculating the density profile and through a calculation of the chemical potential. For liquid layers adsorbed onto the inert solids, we find self-bound liquid phases in both the third and fourth layers and determine the equilibrium density. In the third layer at coverages below equilibrium, we find liquid droplets and a metastable uniform liquid phase, and determine the spinodal point that separates these regions. The above phases and their coverage ranges are in good agreement with several experiments. The superfluid density as a function of coverage is also calculated, and it is observed to change only weakly around the promotion density. For coverages above the beginning of fourth layer promotion, we observe continued increase in the third layer density. We note that the third layer density increase is perhaps enough to cause solidification in this layer, which would explain heat capacity peaks observed experimentally for fourth layer coverages and would provide a simple explanation for the plateaus seen in the superfluid coverage. For helium adsorbed on an active second layer, we observe that a self-bound liquid phase occurs in the third layer, and we determine the equilibrium density and spinodal point, which remain in agreement with experiment. We find that promotion to both the third and fourth layers is signaled by a change in the density dependence of the chemical potential. We further observe the increase in the second layer density with increasing total coverage. The coverage dependence of the superfluid density is calculated, and a pronounced drop is seen at high third layer coverages, as has also been observed experimentally.

Local superfluidity in inhomogeneous quantum fluids

“…23 we introduced an approximate local superfluid estimator that was based on the exchange length of Feynman paths. Using this estimator, local decomposition of superfluid and nonsuperfluid densities in the first solvation shell around the molecule yielded rotational constants of SF 6 and OCS molecules inside the helium droplets that are in good agreement with their experimentally measured values. 10,23 However, this exchange-length-based superfluid estimator was only qualitative and did not provide a rigorous decomposition of the global superfluid response.…”

Path integral Monte Carlo applications to quantum fluids in confined geometries

“…This experimental fact is also seen in the PIMC simulation of PM. While the simulation 33 indicates that third layer remains liquid at completion, an increased spatial order is observed indicating that the layer at completion is near solidification. While the coverage is continually increased and the fourth layer liquid is built, atoms are observed to go to the third layer.…”

“…The growth of 4 He on graphite can be experimentally studied layer by layer and the coverage per unit surface area for layer promotion can be accurately determined. 28 Pierce and Manousakis ͑PM͒ [29][30][31][32][33] using PIMC, the Carlos-Cole 34 helium-graphite interaction and the Aziz helium-helium potential have found that these promotion coverages can be reproduced within a few percent up to the fifth layer. The first layer phase diagram as a function of coverage and temperature has a ͱ3ϫͱ3 commensurate solid at 1/3 coverage.…”

Liquid and Solid Phases ofHe3on Graphite

“…[14], it is shown that the possibility of 3 He atoms moving perpendicularly to the surface leads to a stable liquid phase when the substrate is weakly attractive, as on some alkali metal surfaces. This is probably expected because the system goes from a 2D film to a three-dimensional (3D) configuration where liquid 3 He is the ground-state phase.In this work, we concerned ourselves with the adsorption of 3 He on a clean surface of graphite, trying to reproduce the recent experimental findings of Sato et al[1] Our goal was to bridge the discrepancy between the strictly 2D calculations and the experimental data by improving the theoretical description of the system.Since considering a quasi-two dimensional flat adsorbent is clearly not enough for graphite [14], we included the effects of the substrate corrugation on the behavior of the adsorbate, in line with what has been done previously for 4 He on the same system [15][16][17][18][19][20][21][22]. We found that a corrugated surface is the missing ingredient to reconcile the experimental and theoretical data.…”

“…Since considering a quasi-two dimensional flat adsorbent is clearly not enough for graphite [14], we included the effects of the substrate corrugation on the behavior of the adsorbate, in line with what has been done previously for 4 He on the same system [15][16][17][18][19][20][21][22]. We found that a corrugated surface is the missing ingredient to reconcile the experimental and theoretical data.…”