The structure of quantized vortices in superfluid ³He-A in a large field (>> dipolar field) could be deciphered on the basis of a small, well-shifted satellite peak in the NMR spectrum [PRL 49, 1258 (1982)]. The satellite peak can be explained only by doubly quantized, continuous vortices. The core size for a 4π vortex without a singularity is large enough to localize spin waves that have a sufficiently large frequency shift from the continuum in order to fit the experimental observations.

The NMR satellite in A-phase is observed in addition to the Ω-dependent broadening of the main peak observed in the first experiments. This satellite remained unobserved in our previous experiments because its amplitude is very small.

The vortex satellite was observed only during rotation; its frequency shift and intensity behaved reproducibly. Because this satellite is broad, it can be seen as a separate peak only sufficiently far from Tc, mainly in supercooled He3-A, where the bulk A-phase frequency shift is large. The area of the satellite depended linearily on the angular velocity as expected for vortices, the number of which is linear in Ω. This linear dependence, combined with the result that the satellite frequency does not depend on Ω, provided compelling evidence for vortices in rotating He3-A. Due to the strength of the satellite peak, it was concluded that the satellite was produced by localized spinwave modes trapped on each vortex core.

FIGURE Transverse NMR absorption line shapes at 1—T/Tc = 0.267 as a function of frequency showing the satellite peaks as well as the Ω-dependent broadening of the main A-phase peak. Part (b) is a closeup of part (a), showing the soliton peak at Ω = 0 and the vortex peak at Ω = 1.21 rad/s. In the latter case the soliton peak has already decayed.