Lefschetz fibrations and symplectic homology (Q1016792)

The paper under review is to prove that there are infinitely many finite type Stein manifolds diffeomorphic to each other, but pairwise distinct as symplectic manifolds. A symplectic manifold \((V, \omega)\) is called Stein if there exists a complex structure \(J\) and a proper and bounded from below plurisubharmonic function \(\phi : V \to \mathbb R\) such that \(\omega = - d d^c\phi \) and \(d^c(a)(X) = da (JX)\). The triple \((V, J, \phi)\) is a Stein manifold. If \((J_t, \phi_t)\) is a smooth family of Stein structures on \(V\), then it is called a Stein deformation if \(\phi : (t, x) \mapsto \phi_t(x)\) is proper and for each \(t\in [0, 1]\) there exists a sequence \(\{c_k\}_{1\leq k \leq \infty}\) such that \(c_k < c_{k+1}\), \(\lim_{k\to \infty}c_k = \infty\) and\(c_k\) is a regular value of \(\phi_s\) for \(s\in (t-\varepsilon, t+\varepsilon)\). Two Stein manifolds are equivalent if they are symplectically diffeomorphic to each other or they are Stein deformation equivalent. The proof is based on a construction of Stein manifolds. Let \(V=\{x^7+y^2+z^2+w^2=0\}\subset \mathbb C^4\) and \(p=(0,0,1,i)\in V\). Let \(H\) be a blowup of \(\mathbb C^4\) at \(p\). Then \(X = H \smallsetminus \tilde{V}\) is a Stein manifold which can be seen as (1) cut out the hypersurface \(V\) in \(\mathbb C^4\) to get \(Z = \mathbb C^4 \smallsetminus V\) and (2) blow up \(Z\) at infinity to get \(X\). \(X\) is called the Kaliman modification of \((\mathbb C^4, V, p)\). One can consider the end connected sum \(X_n = \#_{i=1}^n X\). Two Stein manifolds with \(H_1=0\) have the same symplectic homology. The symplectic homology of the end connected sum is the product of the symplectic homologies of each component. The invariant used in the proof is the number \(i(X)\) of idempotents of symplectic homology. This reduces to show that \(1< i(X) < \infty\) and \(i(X_n) = i(X)^n\), and the ring structure of the symplectic homology of \(X\) is isomorphic to the ring structure of the symplectic homology of \(Z\), where \(Z\) is so much easier than \(X\). This is the heart of the main result given in Theorem 1.3. First, the author constructs a \(\mathbb Z_2\)-graded algebra related to a Lefschetz fibration \(E\) and shows that it is equal to symplectic homology in section 4 and section 5.1. In section 6, the author proved that both Lefschetz fibrations have isomorphic \(\mathbb Z_2\)-graded algebra. This will lead to show that \(i(X) < \infty\) in Theorem 7.2. Hence Theorem 1.1, stated as there exists a family of finite type Stein manifolds \(X_i\) (diffeomorphic to \(\mathbb R^{2k}, k\geq 4\)) such that \(i\neq j \in N\) implies \(X_i\) is not equivalent to \(X_j\), is proved. There is no analogue in dimension 4 since any finite type Stein manifold diffoemorphic to \(\mathbb R^4\) is symplectically diffeomorphic to \(\mathbb R^4\). Section 2 give a fair background on Stein manifold, Lefschetz fibration, symplectic homology and Kaliman modification. Section 3 provides the proof of Theorem 1.1. Appendixes are also helpful in understanding the details inside the paper. Note that Gompf constructed uncountably many nonfinite type Stein manifolds which are homeomorphic to \(\mathbb R^4\), but are not pairwise diffeomorphic to each other. It would be interesting to know if there is also a uncountably many finite type exotic Stein manifolds. Also the dimension 6 case for finite type Stein manifolds remains open.