Hypercontact structures and Floer homology (Q834389)

The paper under review introduces a Floer type theory for a pair consisting of Cartan hypercontact 3-manifold and a hyper-Kähler manifold \(X\). The theory is based on the Morse-Floer type of a hyper-symplectic action and its critical solutions as well as indexes and gradient flows. \textit{H. Geiges} and \textit{J. Gonzalo} introduced in [Invent. Math. 121, No.~1, 147--209 (1995; Zbl 1002.53501)] hypercontact structures and proved that every Cartan hypercontact 3-manifold is diffeomorphic to a quotient of the 3-sphere by a right action of a finite subgroup of \(Sp(1)\). A hypercontact structure on a 3-manifold \(M\) is a triple of contact forms \(\alpha =(\alpha_1, \alpha_2, \alpha_3)\in \Omega^1(M, \mathbb R^3)\) such that \[ \alpha_i \wedge d\alpha_j = \delta_{ij} \sigma, \;\;\;i, j =1, 2, 3, \] where \(\sigma\) is a volume form of \(M\). The Reeb vector fields \(v_1, v_2, v_3\) are positively independent and preserve the volume form \(\sigma\). The hypercontact structure is positive if they form a positive frame of the tangent bundle. A positive hypercontact structure is Cartan if \(\alpha_i (v_j) = \delta_{ij}\). Define a natural 1-form on the space of smooth functions \(f: M \to X\) to be \[ A(f)=-\int_M \sum_{i=1}^3 \alpha_i \wedge f^*\omega_i, \] where \(X\) is a hyper-Kähler manifold with complex structure \(I, J, K\) with associated symplectic forms \(\omega_1, \omega_2, \omega_3\). The 1-form \(A\) is closed since those Reeb fields are volume preserving. The zeros of \(A\) are solutions of the nonlinear elliptic first-order partial differential equation \[ \partial_{I, J, K}f=I \partial_{v_1}f + J \partial_{v_2}f + K \partial_{v_3}f = 0. \] The authors prove an existence result for the solutions of the perturbed nonlinear Dirac equation \(\partial_{I, J, K}(f, H) = \partial_{I, J, K}f - \nabla H(f) = 0\) when the linearized operator is self-adjoint. The space of solutions is compact for compact Cartan hypercontact 3-manifolds and 3-tori and a compact hyper-Kähler manifold \(X\), and the number of contractible nondegenerated solutions is bounded below by the sum of the \(\mathbb Z_2\)-Betti numbers of \(X\). The Floer chain complex is generated by the perturbed solutions and the boundary map is given by the finite energy solution \(u: M \to\mathbb R\to X\) of the gradient flow equation \(\partial_s u = - \partial_{I, J, K}(f, H)\). The main technical difficulty is the compactness. Section 3 of the paper is devoted to the regularity and compactness with a key energy identity. Theorem 3.6 gives the compactness of the critical points for the perturbed solutions of \(\partial_{I, J, K}(f, H) =0\), and Theorem 3.13 and Theorem 3.15 give the exponential decay for nondegenerate solutions and the compactness of the moduli space of gradient flows. Section 4 gives the transversality and Fredholm property with index. Theorem B is proved in Section 5 showing that the Floer chain complex is independent of the perturbed function \(H\) and is naturally isomorphic to \(H_*(X, Z_2)\) for the component \(\tau_0\) of constant maps. The paper proposes a Floer-Donaldson theory for this setup and gives a conjectural correspondence between the Donaldson-Thomas-Floer theory of \(M\times S\) with the hypercontact Floer cohomology \(HF^*(M, {\mathcal M}(S))\), where \(S\) is a hyper-Kähler surface and \({\mathcal M}(S)\) is a suitable moduli space of bundles over \(S\). It is quite interesting to code the Donaldson-Thomas theory with a complexified Floer theory.