No abstract

Schrenk, König, and Pobell Reply: In their Comment, Adams et al. [1] claim that the results on the nuclear magnetic ordering of 3 He clusters in a solid 4 He matrix reported by Schrenk et al. [2] can be understood as a result of surface nucleation, hysteresis between melting and freezing, and incomplete melting of 3 He in confined geometries. In this Reply we will show that the fact that we have simultaneously monitored the pressure of the sample while taking heat capacity data [2] provides clear experimental evidence that the interpretation given by Adams et al. can be ruled out as a possible explanation of our results. We want to emphasize that the origin for the existence of solid 3 He in the droplets at pressures far below the bulk 3 He melting curve is still unknown to us.Adams et al. argue that 3 He in the droplets nucleates on the 4 He surface similar to the nucleation of 4 He on Grafoil [3] or 3 He on Grafoil precoated with 4 He [4], leading to a decreasing density gradient in the 3 He droplets from the 4 He-3 He interface to the core of the droplets where at a pressure below the bulk 3 He melting curve 3 He should then be in the liquid state. Consequently, this density gradient has to result in a continuous freezing of 3 He during the warm-up of the sample from the minimum temperature to the bulk 3 He melting curve which, however, was not observed in our measurements. In our experiment, freezing (or melting) of 3 He in the droplets can be detected to a high accuracy by measuring the pressure changes in the sample. This enables us to monitor pressure changes of Dp 0.3 mbar corresponding to the melting (or freezing) of about 0.1% of the total amount of 3 He in the droplets.We have observed at, e.g., p 33.6 bars a constant pressure (to within our experimental resolution) during the warm-up of the sample at temperatures below ϳ8 mK. At this temperature a significant decrease in pressure of about 30 mbar indicates the solidification of the liquid fraction of the 3 He in the droplets (see Fig. 2 in Ref.[2]). Subsequent cooling of the sample again shows that the same amount of 3 He melts without any hysteresis between freezing and melting. We observe a shift of the melting (or freezing) curve of the liquid part of the separated 3 He towards lower temperatures compared with bulk 3 He, but there is no sign of density gradients in the 3 He droplets as expected from substrate nucleation.We can apply the same arguments to exclude the explanation for the history-dependent ordering temperature given by Adams et al. [1]. They explained the history dependence of the specific heat with hysteresis in melting and freezing of the 3 He separated in the droplets. As already mentioned above, melting (freezing) of 3 He results in a pressure increase (decrease which, however, was not observed during the investigation of the history-dependent part of the specific heat. Moreover, at p 36.4 bars, i.e. at a pressure which is 2 bars above the bulk 3 He melting curve and at which clearly only solid 3 He should be present in the dropl...

We have investigated the specific heat of liquid 3 He confined to an Ag sinter with an average pore size of about 1000 Å in the temperature range 1 mKрTр20 mK and at pressures 4.8 barр pр34.0 bar. The specific heat of normal-fluid 3 He in the sinter pores shows the linear temperature dependence expected for a Fermi liquid. However, the effective mass of the 3 He quasiparticles is clearly enhanced in the restricted geometry compared to data obtained in bulk 3 He. In addition, there is a temperature-independent contribution to the specific heat, the origin of which can be interpreted as the specific heat of the second layer of 3 He on the Ag surface. Moreover, compared with the results obtained for bulk 3 He, we observe a much broader maximum in the specific heat in the vicinity of the superfluid transition; this maximum occurs about 0.4 mK below the bulk superfluid transition temperature. Furthermore, in the confinement of the sinter only a part of the 3 He in the sinter ͑about 60%͒ becomes superfluid. In contrast to the results obtained with pure 3 He the specific heat of a liquid 3 He-4 He mixture ͑1% 3 He͒ in the Ag sinter shows no deviation from bulk data.

No abstract

No abstract