Surgery obstructions and Heegaard Floer homology (Q315270)

The article under review deals with the problem of obstructing an integer homology sphere \(Y\) from being surgery on a knot. If \(Y\) is reducible, then a result of \textit{C. McA. Gordon} and \textit{J. Luecke} [J. Am. Math. Soc. 2, No. 2, 371--415 (1989; Zbl 0678.57005)] implies that it can never be surgery on a knot. If \(Y\) is irreducible, it can be toroidal, hyperbolic or Seifert fibred. \textit{D. Auckly} [AMS/IP Stud. Adv. Math. 2, 21--34 (1997; Zbl 0889.57022)] gave examples of toroidal and hyperbolic integer homology spheres which are not surgery on a knot. The article under review covers the last case. First, the authors prove an obstruction to integer homology spheres from being surgery on a knot (Theorem 1.2) in terms of Heegaard Floer homology: If \(Y=S^3_{1/n}(K)\) for some integer \(n\) and some knot \(K \subset S^3\), such that \(d(Y)\leq{-8}\), then \(U\cdot\mathrm{HF}_0^{\mathrm{red}}(Y) \neq0\), where \(\mathrm{HF}_0^{\mathrm{red}}(Y)\) denotes the reduced Heegaard Floer homology of \(Y\) in grading \(0\). This is established by using Ozsváth and Szabó's description of the Heegaard Floer homology of a surgery as a mapping cone [\textit{P. S. Ozsváth} and \textit{Z. Szabó}, Algebr. Geom. Topol. 11, No. 1, 1--68 (2011; Zbl 1226.57044)], and by chasing elements in the diagram. Then, the authors show that \(Y_p = \Sigma(p, 2p-1, 2p+1)\) satisfies \(d(Y_p)\leq{-8}\) (for \(p\) large enough), and \(U\cdot\mathrm{HF}_0^{\mathrm{red}}(Y_p) = 0\), and therefore it is not surgery on a knot in \(S^3\). The authors also note that a classical obstruction (weight one fundamental group) fails to detect that \(Y_p\) is not a surgery. The computation of \(d(Y_p)\) follows from the fact that \(Y_p\) is the boundary of a plumbing. The action of \(U\) on \(\mathrm{HF}_0^{\mathrm{red}}(Y_p)\) is determined by using a result of \textit{M. B. Can} and \textit{Ç. Karakurt} [Acta Math. Hung. 144, No. 1, 43--75 (2014; Zbl 1340.57010)], which gives a combinatorial description of the \(\mathrm{HF}^+(\Sigma(p,q,r))\), in terms of ``graded roots''. A large part of the article under review is devoted to unpacking the graded root description, and to determining part of \(\mathrm{HF}^+(Y_p)\) explicitly. Lastly, the authors discuss further applications of their techniques to surgeries on knots in 3-manifolds other than \(S^3\).