Supersolid He not so super after all?

For the past couple of years I was one of those people who are quite fascinated with the story of supersolid He. It's one of those things where the physics is really intriguing and beautiful, but experiments are relatively easy to understand, even though they are technically quite challenging.

Basic idea: if you take a bucket, fill it with water and spin it around its axis, water will contribute to moment of inertia (the technique is called "torsional oscillator" and is similar to pendulum, except one uses rotation instead of lateral motion). If instead of water you use He and cool the system down, at some point the contribution from He vanishes as it becomes superfluid (through a process known as Bose-Einstein condensation). This is a well-known phenomena - but what Moses Chan and group at Penn State have discovered in 2004 is that the same effect can be seen for solid He, which can be obtained by applying a relatively modest pressure at low temperatures.

One possible and exciting explanation is that solid He "flows" without resistance, therefore becoming "super"-solid.

The results were somewhat controversial - for a while everyone was waiting for other groups to reproduce the effect, and a lot of people talked about Moses getting a Nobel Prize for this work. The findings of Chan were then reproduced by other groups, but details were a bit murky - some groups reported that annealing of crystals made the effect go away, implying that disorder plays a crucial role. This contradicted Chan's claims that the annealing either did not have much of an effect or made things worse.

To give you an idea of difficulty of such experiments - only about 2% of solid He turns "supersolid", and if you adjust for overall torsional oscillator setup, it corresponds to 0.4% drop in moment of inertia. The effect is most pronounced at temperatures below 100mK and requires pressures > 25 bar to get He into solid phase. Solid He has consistency comparable to that of butter at these conditions, so it is not solid in the way we think of most rocks or metals.

The role of defects, such as grain boundaries, is a crucial - as pointed out by several people, most materials pre-melt, that is form a fluid-like layer coating the surface of a crystal well below bulk melting point. This surface melting is the reason why we can enjoy skating on ice, skiing and other wintertime activities (and not pressure-induced lowering of melting point, as is erroneously taught sometimes). Surface melting occurs because molecules at the boundaries of a crystal are more prone to disorder.

So could supersolid behavior be due to the pre-melting at the grain boundaries, and flowing of the superfluid component, rather than true supersolid effect?

Chan and coworkers attempted to answer this by confining He in a vycor glass - a randomly porous matrix, which will effectively increase the amount of disorder by increasing surface-to-volume ratio - and they have not seen much change in the observed supersolid fraction.

On the other hand, it was always puzzling why there was no pressure-induced flow that one would expect, and only torsional oscillator experiment yielded positive results.

So the whole state of affairs was quite messy and confusing for a while.

For more details go to KITP collection of talks on this very topic from early 2006.
As you can see, a talk by Greg Dash, an expert on pre-melting, casts some doubt on Chan's results by arguing that this could be a rather benign and well-known premelting at grain boundaries.

Later last year some theorists added more fuel to the fire by claiming that theoretically one would not expect to see the supersolid effects under conditions where it was seen.

And now a report in Science from recent APS meeting in Denver claims scientists have reached a consensus that the effect is indeed due to flow at grain boundaries, and not supersolidity.
It was a little strange to see theorists getting lions share of credit - from where I sit it seemed like experimentalists who did annealing experiments, including Chan himself but also groups in France (Balibar) and Canada, in addition to premelting experts like Dash who I thought should get most of the credit for bringing consensus - especially as theorists could not agree on what the real effect - for example a recent paper in Nature Physics by Phil Anderson claims that supersolidity in He is vortex fluid effect.