It has been proposed that the observed nonclassical rotational inertia (NCRI) in solid helium results from the superflow of thin liquid films along interconnected grain boundaries within the sample. We have carried out new torsional oscillator measurements on large helium crystals grown under constant temperature and pressure. We observe NCRI in all samples, indicating that the phenomenon cannot be explained by a superfluid film flowing along grain boundaries.

We have investigated the influence of impurities on the possible supersolid transition by systematically enriching isotopically-pure 4 He (< 1 ppb of 3 He) with 3 He. The onset of nonclassical rotational inertia is broadened and shifts monotonically to higher temperature with increasing 3 He concentration, suggesting that the phenomenon is correlated to the condensation of 3 He atoms onto the dislocation network in solid 4 He.

No abstract

We have carried out torsional oscillator (TO) and heat capacity (HC) measurements on solid 4 He samples grown within a geometry which restricts the helium to thin (150 micron) cylindrical discs. In contrast to previously reported values from Rittner and Reppy of 20% non-classical rotational inertia (NCRI) for similar confining dimensions, 0.9% NCRI (consistent with that found in bulk samples and samples imbedded in porous media) was observed in our TO cell. In this confined geometry the heat capacity peak is consistent with that found in bulk solid samples of high crystalline quality.PACS A tremendous amount of work has been done in an attempt to understand the microscopic origin of the NCRI signals [7][8][9]. These signals have been reproduced in many laboratories [10][11][12][13][14][15][16]. While the temperature dependence of NCRI is well reproduced, the magnitude of NCRI is highly sample-dependent. In the experiment of Kim and Chan where the samples were confined within an annulus with a width of ~1 mm, NCRI was found to vary between 0.5% and 1.4% [2]. Many of the TO measurements were on bulk solid 4 He grown in cells with cylindrical sample spaces with typical dimensions of ~1 cm [17]. The NCRI fraction (NCRIF) in these cylindrical samples varies between 0.015% and ~1%. In some samples the NCRIF can be reduced, sometimes dramatically, by thermal annealing [17,18]. The annealing effect and the variation in the NCRIF support the interpretation that various forms of disorder in the solid (grain boundaries, dislocation lines, glassy phases [19][20][21][22]) are either responsible for, or at least enhance the NCRI signal. Nevertheless, experiments on solid helium confined within disordered porous media such as Vycor glass [1] and porous gold [23] (with confining dimensions of 7 nm and 500 nm respectively) also found NCRI only on the order of 1-2%. A recent attempt to induce disorder in the helium by growing the solid inside of 95% porous silica aerogel did not lead to an enhanced NCRIF [24]. Rittner and Reppy (RR) carried out a series of TO measurements with the solid samples confined to thin annuli with 300, 150 and 74 µm gaps. Remarkably, they observed NCRIF as high as ~20% [18,25]. The results of RR suggest that there is an optimal surface to volume ratio, S/V, of the solid helium sample that is on the order of 100 to 1000 cm -1 (or with confining dimension of ~150 µm) that is particularly conducive for the appearance of large NCRI. This extraordinary large NCRI is one of the most intriguing and puzzling results in the subject of superflow in solid helium.In this paper we report a TO study of solid helium with confining dimension and S/V ratio similar to RR (~400 cm -1 ). The internal geometry of the cell is similar to the original Andronikashvili liquid helium experiment [26]. The helium is confined to the space between thinly separated discs (150 µm gap). The oscillation axis is perpendicular to the plane of the thin discs. In contrast to RR, the NCRI observed in our sandwiched samples is less than 1%, in ...

We have studied natural purity 4 He single crystals and polycrystals between 10 and 600 mK using a torsional oscillator with a 2 cm 3 rigid cell made of sapphire with a smooth geometry. As the temperature was lowered, we observed sample dependent but reproducible resonant frequency shifts that could be attributed to a supersolid fraction of order 0.1%. However, these shifts were observed with single crystals, not with polycrystals. Our results indicate that, in our case, the rotational anomaly of solid helium is more likely due to a change in stiffness than to supersolidity. This interpretation would presumably require gliding of dislocations in more directions than previously thought.