On manifolds homeomorphic to \({\mathbb{C}}P^ 2\#8\overline{{\mathbb{C}}P}^ 2\) (Q582636): Difference between revisions

The following problem has been of considerable interest recently: Which is the smallest value of k (denote it by \(k_ 0)\) such that there is a complex surface homeomorphic but not diffeomorphic to \({\mathbb{C}}P^ 2\#k\overline{{\mathbb{C}}P}^ 2\), equipped with the standard smooth structure? It was shown earlier by \textit{R. Friedman} and \textit{J. W. Morgan} [J. Differ. Geom. 27, 297-369 (1988; Zbl 0669.57016)] and, independently, by \textit{C. Okonek} and \textit{A. Van de Ven} [Invent. Math. 86, 357-370 (1986; Zbl 0613.14018)] that \(k_ 0\leq 9\). The main result of the paper under review is that \(k_ 0\leq 8\). Namely, it is demonstrated that the Barlow surface is not diffeomorphic to \({\mathbb{C}}P^ 2\#8\overline{{\mathbb{C}}P}^ 2\) (although these two 4-manifolds are homeomorphic by M. H. Freedman's classification theorem). A key ingredient of the proof is a new invariant \(\phi\) defined for oriented Riemannian 4-manifolds X such that \(b_ 1(X)=0\), \(b^+_ 2(X)=1\), and \(b^-_ 2(X)=8k\), which is simpler than S. K. Donaldson's \(\Gamma_ X\)-invariant and polynomial invariants.