Heegaard Floer homology and cosmetic surgeries in \(S^3\) (Q6172664)

Given a 3-manifold and a knot, Dehn surgery is an operation to make a ``new'' 3-manifold as follows: Remove an open tubular neighborhood of the knot, and glue a solid torus back. It is naturally expected that a non-trivial Dehn surgery gives a new manifold, that is, the 3-manifold so obtained is not homeomorphic to the original manifold. This is equivalent to the fact that no two inequivalent Dehn surgeries give the same (orientation-preservingly homeomorphic) 3-manifolds. This plausible claim, which was originally conjectured in the affirmative by \textit{C. McA. Gordon} [in: Proceedings of the international congress of mathematicians (ICM), August 21-29, 1990, Kyoto, Japan. Volume I. Tokyo etc.: Springer-Verlag. 631--642 (1991; Zbl 0743.57008)], has not been confirmed yet. There are many positive evidences by using several invariants of 3-manifolds and their surgery formulae. In the paper under review, by making use of the surgery formula for Heegaard Floer homology, the author shows that if a knot in the 3-sphere admits a pair of truly cosmetic surgeries, that is, a pair of Dehn surgeries yielding orientation-preservingly homeomorphic 3-manifolds, then the surgery slopes are either 2 or \(1/q\) for some value of \(q\) that is explicitly determined by the knot Floer homology of the knot. Moreover, in the former case the genus of the knot must be 2, and in the latter case there is a bound relating \(q\) to the genus and the Heegaard Floer thickness of the knot. In particular, this implies that for any given knot, there can be at most two pairs of truly cosmetic surgeries. The key advantage of this paper is to use a recent reinterpretation of Heegaard Floer invariants in terms of collections of immersed curves obtained by the author together with Rasmussen, and Watson [\textit{J. Hanselman} et al., ``Bordered Floer homology for manifolds with torus boundary via immersed curves'', Preprint, \url{arXiv:1604.03466} and Proc. Lond. Math. Soc. (3) 125, No. 4, 879--967 (2022; Zbl 1530.57018)]. This facilitates getting stronger obstructions than previously known by considering the full graded theory.