A note on positive-definite, symplectic four-manifolds (Q1757302)

If signatures and Euler characteristics can be realized by closed, symplectic four-manifolds, then this is called the geography problem for symplectic four-manifolds. One of the purposes of the geography problem is to understand the difference between symplectic and complex or Kähler manifolds. The example illustrating this is the Bogomolov-Miyaoka-Yau inequality stating that every complex surface \(S\) of general type satisfies \(3\sigma(S)\leq\chi(S)\), where \(\sigma(S)\) is the signature and \(\chi(S)\) is the characteristic of \(S\). It is an open question whether symplectic manifolds satisfy this as well. In this paper, the authors consider the case where the manifold \(X\) is positive-definite with \(b^+_2\geq 2\). Then \(X\) is homeomorphic to \(\#_n\mathbb CP^2\) which is not symplectic for any \(n\geq 2\), and therefore, a simply-connected, positive-definite, symplectic four-manifold would be an exotic \(\#_n\mathbb CP^2\), for which there are currently none known. A manifold \(X\) is called geometrically simply-connected if it admits a handlebody decomposition with no 1-handles which is equivalent to having a decomposition with no 3-handles instead and such a manifold is simply-connected. The goal of this paper is to prove that under mild conditions, simply-connected, positive-definite, symplectic manifolds do not exist. The authors show that if \(X\) is a closed, geometrically simply-connected four-manifold with \(b_2^+\geq 2\), and has either no 1-handles or no 3-handles, then \(X\) is not symplectic. In this case, \(X\) satisfies \(\sigma(X)+4\leq\chi(X)\). If the trace \(W_n(K)\) of an integral surgery \(S^3_n(K)\) is the four-manifold obtained by attaching an \(n\)-framed 2-handle to the four-ball along a knot \(K\), then the authors show that if \(K\) in \(S^3\) satisfies \(d(S^3_1(K))=0\), then the trace \(W_n(K)\) cannot be a symplectic filling of \(S^3_n(K)\) for \(n>0\) and any contact structure on \(S^3_n(K)\).