A POSSIBLE EXPLANATION FOR THE RECENTLY OBSERVED PHASE TRANSITION IN SOLID 3He IN HIGH MAGNETIC FIELDS

Schuberth, Bakalyar and Adams (SBA) [1] have supercurrents may imply that magnetization can be recently reported evidence for a phase transition on supplied rapidly enough to allow solid to form with the 3He melting curve near the A 1 —A2 superfluid tranequilibrium polarization. Using the following expressitions in high magnetic fields. The phenomenon was sion for the entropy of the solid [5]: manifested as a backstep in the chart traces of pressure ~ S(T H)s(T 0) 1 / pH \2 versus time. There were, however, several odd charac= ‘ Nk ‘ = —~ kk(T— 9 ) teristics. For example, although the feature has only been seen in high magnetic fields (—‘20 kG), it occurs where 0 is the Néel temperature, we estimate z~S/Nk at the same temperature for different fields. In addi—‘0.1 in 2 in agreementwith latent heat measurements. tion, it was not observed when the sample was warmed. This is a larger drop in entropy than that associated Latent heat measurements indicated that there was a with the pressure versus time chart traces because of significant drop in the entropy (‘~-‘0. lRln 2). It should solid formation is much faster in the latent heat meabe kept in mind that the magnetic properties of liquid surements. 3He change dramatically as the temperatur decreases We can also compare the rate of change of the magbelow the A superfluid transition temperature Tc. netization of the solid and the liquid above and below Corruccini and Osheroff [2] have found that the longiTc. The change in T 1 at the superfluid transition mditudinal spin relaxation time T1 drops by three orders catesthat the spin relaxation mechanism has changed. of magnitude below Tc. Anderson 13] and Vuorio [4] Amore specific model involves the onset of supercurhave suggested that this drop can be explainedby magrents which are much more effective in transporting netic supercurrents of spin-up and spin-down supermagnetization than is spin diffusion. We now present fluids. We propose the following explanation of the two calculations—one with T1 and one with spin difphenomenon seen by SBA. Above the superfluid tranfusion and supercurrents causing the change in the magsition, the solid that is formed by compression may netization deficit. not have equilibrium magnetization because spin reIn order to form solid withequilibrium magnetizalaxation and spin diffusion processes are too slowto tion, the rate at which the magnetization must be supprovide the necessary magnetization to the normal plied to the solid can be expressed as liquid from which the solid is formed. This magnetiM5m(dM/dt) rd~(dY/dt)x H, zation deficit mcreases with the rate of compression. S S