Models for Superfluid 3He in Aerogel

Aerogel is a very porous material where, for example, 98% of the volume can be empty. The figure represents some strands of aerogel. The dimension of the figure corresponds to roughly 100 nm. More about aerogels: Los Alamos and Berkeley.

We are interested in the case that liquid 3He fills the aerogel at very low temperatures. Pure 3He goes into superfluid state at temperatures 1 ... 3 mK, depending on pressure. When 3He is inside aerogel, the superfluid transition is suppressed. Experiments studying superfluid 3He in aerogel are made recently in Cornell University, Northwestern University, University of Manchester, and Stanford University.

The main effect of the aerogel on 3He arises because the quasiparticles of 3He are scattered from the aerogel strands. The simplest "homogeneous scattering model" replaces the strands with a uniform scattering medium. In other words, the probability for a quasiparticle to be scattered is the same at any location. This is the standard model when considering impurities in superconducting metals (Abrikosov and Gorkov, 1961). We find that this model gives results with the right tendency, but quantitatively it is quite insufficient. See here for computer code to calculate the parameters of this model.

From the failure of the homogeneous scattering model we infer that the inhomogeneity of the scattering is crucial. One possible model of inhomogeneous scattering is to consider 3He between parallel planes. This gives a better fit to the experiments, but there is a problem that this model is not consistent with the observed isotropy of the aerogel.

Apparently, a right model should allow some random voids in the scattering medium. This would be rather tedious to calculate. Therefore we prefer to study a periodic lattice of voids. Moreover, we approximate the unit cell of this lattice by a sphere. This "isotropic inhomogeneous scattering model" can rather well account for the measured suppression of the transition temperature. It is also superior to the homogeneous model in explaining the measured pairing amplitude and superfluid density. We find that aerogel is full of voids whose radius is on the order of 100 nm. This is roughly consistent with other measurements of the structure of aerogel.

The aerogel is anisotropic on a short length scale. The anisotropy forms effectively a random field on the order parameter of the superfluid. Depending on parameters, the anisotropy can strongly affect the NMR properties. The sudden change in the NMR frequency shift at tipping angles near 50 degrees might be interpreted as a transition from an ordered to a disordered texture.


Back to helium theory page

22.4.2007, Erkki Thuneberg, Email